{ "snapshot": "2026-10-25", "generated_at": "2026-10-26T19:38:25.408Z", "counts": { "predictions": 240, "people": 209, "sources": 127, "first_party_share": 0.4643, "revisions": 3, "verified": 56, "editorial": 184, "graded": 20, "open": 36, "unrated": 184, "government": 35 }, "people": [ { "id": "altman", "name": "Sam Altman", "role": "CEO and cofounder", "org": "OpenAI", "orgShort": "OpenAI", "group": "frontier-labs", "emblem": "SA", "accent": "#d6ff3f", "bio": "Sets OpenAI strategy, infrastructure partnerships and the public case for deployment. Publishes personal essays that move the whole industry’s timeline.", "rosterAnchor": "OpenAI — research, deployment, and nonprofit control", "status": "verified" }, { "id": "amodei", "name": "Dario Amodei", "role": "Cofounder and CEO", "org": "Anthropic", "orgShort": "Anthropic", "group": "frontier-labs", "emblem": "DA", "accent": "#ffb347", "bio": "Writes unusually explicit timelines for “powerful AI” and unusually explicit warnings about what it does to entry-level work.", "rosterAnchor": "Anthropic — Claude, alignment, interpretability, and public benefit", "status": "verified" }, { "id": "hassabis", "name": "Demis Hassabis", "role": "Chair, Google DeepMind; Alphabet chief scientist", "org": "Google DeepMind", "orgShort": "DeepMind", "group": "frontier-labs", "emblem": "DH", "accent": "#7fd7ff", "bio": "Nobel laureate, Isomorphic Labs lead. Has publicly shortened his AGI window at least four times since 2024.", "rosterAnchor": "Google DeepMind and Alphabet", "status": "verified" }, { "id": "pichai", "name": "Sundar Pichai", "role": "CEO", "org": "Google and Alphabet", "orgShort": "Alphabet", "group": "frontier-labs", "emblem": "SP", "accent": "#9dff8f", "bio": "Allocates capital across Gemini, search, cloud and TPU capacity — the widest single distribution surface for AI on earth.", "rosterAnchor": "Google DeepMind and Alphabet", "status": "verified" }, { "id": "musk", "name": "Elon Musk", "role": "xAI founder; SpaceX CEO", "org": "xAI / Grok (within SpaceX)", "orgShort": "xAI", "group": "frontier-labs", "emblem": "EM", "accent": "#ff7a59", "bio": "The most prolific dated forecaster in the industry, and therefore the most testable. His dates move; the definitions do not always move with them.", "rosterAnchor": "xAI/Grok — within SpaceX", "status": "verified" }, { "id": "zuckerberg", "name": "Mark Zuckerberg", "role": "Founder, chairman and CEO", "org": "Meta", "orgShort": "Meta", "group": "frontier-labs", "emblem": "MZ", "accent": "#7f9dff", "bio": "Bets on compute scale plus distribution through three billion-person consumer surfaces, with Meta Superintelligence Labs building the models.", "rosterAnchor": "Meta — Superintelligence Labs and FAIR", "status": "verified" }, { "id": "suleyman", "name": "Mustafa Suleyman", "role": "CEO, Microsoft AI", "org": "Microsoft AI", "orgShort": "Microsoft", "group": "frontier-labs", "emblem": "MS", "accent": "#66e0d0", "bio": "Pursues “humanist superintelligence”: capable, deliberately bounded, and explicitly subordinate. Also the most concrete forecaster of AI companionship.", "rosterAnchor": "Microsoft AI and Amazon model development", "status": "verified" }, { "id": "sutskever", "name": "Ilya Sutskever", "role": "Cofounder and CEO", "org": "Safe Superintelligence (SSI)", "orgShort": "SSI", "group": "frontier-labs", "emblem": "IS", "accent": "#c9a6ff", "bio": "Transformer-era coauthor who now argues the age of pure scaling is over — the most consequential technical disagreement inside the field.", "rosterAnchor": "New general-intelligence and research-automation ventures", "status": "verified" }, { "id": "lecun", "name": "Yann LeCun", "role": "Cofounder and executive chair", "org": "AMI Labs", "orgShort": "AMI Labs", "group": "science-robotics", "emblem": "YL", "accent": "#a8ffce", "bio": "Left Meta to build world models on the premise that large language models are a short-lived paradigm.", "rosterAnchor": "World models and physical intelligence", "status": "verified" }, { "id": "feifei", "name": "Fei-Fei Li", "role": "Cofounder and CEO", "org": "World Labs", "orgShort": "World Labs", "group": "science-robotics", "emblem": "FL", "accent": "#ffd166", "bio": "ImageNet’s architect now argues that spatial intelligence — not more text — is AI’s next frontier.", "rosterAnchor": "World models and physical intelligence", "status": "verified" }, { "id": "bengio", "name": "Yoshua Bengio", "role": "Scientific director, LawZero; Université de Montréal professor", "org": "LawZero / Mila", "orgShort": "LawZero", "group": "researchers", "emblem": "YB", "accent": "#b8f2ff", "bio": "Co-chairs the UN Independent International Scientific Panel on AI and chairs the International AI Safety Report. Reads the capability curves rather than the press releases.", "rosterAnchor": "Scientific direction, learning, and evaluation methods", "status": "verified" }, { "id": "hinton", "name": "Geoffrey Hinton", "role": "University Professor Emeritus", "org": "University of Toronto", "orgShort": "U of T", "group": "researchers", "emblem": "GH", "accent": "#ff9ec4", "bio": "Nobel laureate in physics. Has moved his own timelines inward twice in public, and says so.", "rosterAnchor": "Scientific direction, learning, and evaluation methods", "status": "verified" }, { "id": "barnes", "name": "Beth Barnes", "role": "Founder and CEO", "org": "METR", "orgShort": "METR", "group": "researchers", "emblem": "BB", "accent": "#9ff0c0", "bio": "Runs the evaluation organisation whose “time horizon” chart became the industry’s most cited forecast — and the most cited caveat about forecasts.", "rosterAnchor": "Frontier-risk assessment and independent scrutiny", "status": "verified" }, { "id": "leike", "name": "Jan Leike", "role": "Leads the Alignment Science team", "org": "Anthropic", "orgShort": "Anthropic", "group": "researchers", "emblem": "JL", "accent": "#ffcf8a", "bio": "Co-led OpenAI’s Superalignment programme, then left it. The four-year commitment he signed is now a test case in institutional accountability.", "rosterAnchor": "Anthropic — Claude, alignment, interpretability, and public benefit", "status": "verified" }, { "id": "narayanan", "name": "Arvind Narayanan", "role": "Professor; director, Center for Information Technology Policy", "org": "Princeton University", "orgShort": "Princeton", "group": "researchers", "emblem": "AN", "accent": "#ffd9a0", "bio": "Co-author of AI Snake Oil. Separates demonstrated capability from commercial narrative, and argues that x-risk numbers are feelings wearing lab coats.", "rosterAnchor": "Frontier-risk assessment and independent scrutiny", "status": "verified" }, { "id": "ng", "name": "Andrew Ng", "role": "Founder, DeepLearning.AI; managing GP, AI Fund", "org": "DeepLearning.AI", "orgShort": "DeepLearning.AI", "group": "researchers", "emblem": "AN", "accent": "#c6ff7a", "bio": "The industry’s most-read teacher. His forecasts are about adoption and engineering practice rather than raw capability.", "rosterAnchor": "Scientific direction, learning, and evaluation methods", "status": "verified" }, { "id": "gebru", "name": "Timnit Gebru", "role": "Founder and executive director", "org": "Distributed AI Research Institute (DAIR)", "orgShort": "DAIR", "group": "critics", "emblem": "TG", "accent": "#ff8fa3", "bio": "Argues the existential-risk frame is a distraction from harms that are already measurable — and a route to regulatory capture.", "rosterAnchor": "Rights, labor, privacy, and the political economy of technology", "status": "verified" }, { "id": "bender", "name": "Emily M. Bender", "role": "Professor of linguistics", "org": "University of Washington", "orgShort": "UW", "group": "critics", "emblem": "EB", "accent": "#ffb3d1", "bio": "Lead author of “On the Dangers of Stochastic Parrots”. Her objection is about meaning, not benchmark scores.", "rosterAnchor": "Rights, labor, privacy, and the political economy of technology", "status": "verified" }, { "id": "delangue", "name": "Clément Delangue", "role": "Cofounder and CEO", "org": "Hugging Face", "orgShort": "Hugging Face", "group": "infrastructure", "emblem": "CD", "accent": "#ffe066", "bio": "Runs the model-distribution layer, where download counts make open-weight predictions unusually checkable. NVIDIA’s acquisition is pending, not complete.", "rosterAnchor": "Open models and shared software infrastructure", "status": "verified" }, { "id": "huang", "name": "Jensen Huang", "role": "Founder, president and CEO", "org": "NVIDIA", "orgShort": "NVIDIA", "group": "infrastructure", "emblem": "JH", "accent": "#8bff6a", "bio": "Sells the picks for the AI gold rush. His forecasts are simultaneously market analysis and procurement strategy.", "rosterAnchor": "Semiconductors and the physical computing stack", "status": "verified" }, { "id": "su", "name": "Lisa Su", "role": "Chair and CEO", "org": "AMD", "orgShort": "AMD", "group": "infrastructure", "emblem": "LS", "accent": "#ff9f6e", "bio": "The credible second source for AI accelerators. Her total-addressable-market calls have been revised upward twice.", "rosterAnchor": "Semiconductors and the physical computing stack", "status": "verified" }, { "id": "son", "name": "Masayoshi Son", "role": "Chairman and CEO", "org": "SoftBank Group", "orgShort": "SoftBank", "group": "infrastructure", "emblem": "MS", "accent": "#ffd66b", "bio": "The largest single financial bet on artificial superintelligence, denominated in goldfish.", "rosterAnchor": "Cloud capacity, enterprise deployment, devices, and financing", "status": "verified" }, { "id": "gomez", "name": "Aidan Gomez", "role": "Cofounder and CEO", "org": "Cohere", "orgShort": "Cohere", "group": "infrastructure", "emblem": "AG", "accent": "#a5d8ff", "bio": "Attention-paper coauthor building sovereign enterprise AI. Predicts finance is the next white-collar domain to be reorganised.", "rosterAnchor": "Cohere — Canada and enterprise AI", "status": "verified" }, { "id": "robinli", "name": "Robin Li (Li Yanhong)", "role": "Cofounder, chairman and CEO", "org": "Baidu", "orgShort": "Baidu", "group": "infrastructure", "emblem": "RL", "accent": "#ffb3ba", "bio": "Has argued for years that model capability is not where the money is; applications are. China’s most explicit application-layer forecaster.", "rosterAnchor": "China — major platform model developers", "status": "verified" }, { "id": "yang", "name": "Yang Zhilin (Zhilin Yang)", "role": "Cofounder and CEO", "org": "Moonshot AI / Kimi", "orgShort": "Moonshot", "group": "frontier-labs", "emblem": "ZY", "accent": "#bde0fe", "bio": "Built one of China’s most credible frontier challengers and frames agent adoption as a staged curve rather than a switch.", "rosterAnchor": "China — independent model companies", "status": "verified" }, { "id": "brockman", "name": "Greg Brockman", "role": "President and cofounder", "org": "OpenAI", "orgShort": "OpenAI", "group": "frontier-labs", "emblem": "GB", "accent": "#d6ff3f", "bio": "The next decade of progress is decided by whoever can afford to run experiments at scale, not by whoever has the best idea.", "rosterAnchor": "OpenAI — president and cofounder", "status": "editorial" }, { "id": "murati", "name": "Mira Murati", "role": "Founder and CEO", "org": "Thinking Machines Lab", "orgShort": "Thinking Machines Lab", "group": "frontier-labs", "emblem": "MM", "accent": "#ffb347", "bio": "Useful intelligence will be built by teams that tune models to people, not by leaderboards.", "rosterAnchor": "Thinking Machines Lab — founder and ceo", "status": "editorial" }, { "id": "pachocki", "name": "Jakub Pachocki", "role": "Chief scientist", "org": "OpenAI", "orgShort": "OpenAI", "group": "frontier-labs", "emblem": "JP", "accent": "#ff6b6b", "bio": "Capability gains keep compounding from compute and data until the curve itself becomes the strategy.", "rosterAnchor": "OpenAI — chief scientist", "status": "editorial" }, { "id": "mcgrew", "name": "Bob McGrew", "role": "Former chief research officer", "org": "OpenAI", "orgShort": "OpenAI", "group": "frontier-labs", "emblem": "BM", "accent": "#7fd7ff", "bio": "Inference-time reasoning is the next axis of progress, and it will be paid for in silicon.", "rosterAnchor": "OpenAI — former chief research officer", "status": "editorial" }, { "id": "friar", "name": "Sarah Friar", "role": "Chief financial officer", "org": "OpenAI", "orgShort": "OpenAI", "group": "frontier-labs", "emblem": "SF", "accent": "#8bff6a", "bio": "The capital required to stay at the frontier will keep rising faster than revenue, so financing is the real roadmap.", "rosterAnchor": "OpenAI — chief financial officer", "status": "editorial" }, { "id": "simo", "name": "Fidji Simo", "role": "CEO of applications", "org": "OpenAI", "orgShort": "OpenAI", "group": "frontier-labs", "emblem": "FS", "accent": "#c9a6ff", "bio": "AI becomes real when it reaches ordinary workflows, and that means the consumer product is the strategy.", "rosterAnchor": "OpenAI — ceo of applications", "status": "editorial" }, { "id": "weil", "name": "Kevin Weil", "role": "Product lead", "org": "OpenAI", "orgShort": "OpenAI", "group": "frontier-labs", "emblem": "KW", "accent": "#ffd166", "bio": "Every knowledge job will be restructured around a model that drafts first and a human that edits.", "rosterAnchor": "OpenAI — product lead", "status": "editorial" }, { "id": "mchen", "name": "Mark Chen", "role": "Research lead", "org": "OpenAI", "orgShort": "OpenAI", "group": "frontier-labs", "emblem": "MC", "accent": "#9be8c8", "bio": "Models will start producing research results that their own creators cannot fully explain.", "rosterAnchor": "OpenAI — research lead", "status": "editorial" }, { "id": "zaremba", "name": "Wojciech Zaremba", "role": "Cofounder", "org": "OpenAI", "orgShort": "OpenAI", "group": "frontier-labs", "emblem": "WZ", "accent": "#d6ff3f", "bio": "Robotics will follow language models: a general policy trained broadly beats hand-engineered control.", "rosterAnchor": "OpenAI — cofounder", "status": "editorial" }, { "id": "madry", "name": "Aleksander Madry", "role": "Safety researcher", "org": "OpenAI", "orgShort": "OpenAI", "group": "frontier-labs", "emblem": "AM", "accent": "#ffb347", "bio": "Robustness is a research problem, not a compliance problem, and it will be solved or the deployments will not happen.", "rosterAnchor": "OpenAI — safety researcher", "status": "editorial" }, { "id": "olah", "name": "Chris Olah", "role": "Cofounder", "org": "Anthropic", "orgShort": "Anthropic", "group": "frontier-labs", "emblem": "CO", "accent": "#ff6b6b", "bio": "Interpretability is the only way to audit a system nobody can read, and it will become a release requirement.", "rosterAnchor": "Anthropic — cofounder", "status": "editorial" }, { "id": "kaplan", "name": "Jared Kaplan", "role": "Chief science officer", "org": "Anthropic", "orgShort": "Anthropic", "group": "frontier-labs", "emblem": "JK", "accent": "#7fd7ff", "bio": "Scaling laws are the most reliable forecast in the field, and they still hold through the next order of magnitude.", "rosterAnchor": "Anthropic — chief science officer", "status": "editorial" }, { "id": "krieger", "name": "Mike Krieger", "role": "Chief product officer", "org": "Anthropic", "orgShort": "Anthropic", "group": "frontier-labs", "emblem": "MK", "accent": "#8bff6a", "bio": "Agentic products will be judged by what they finish, not by what they answer.", "rosterAnchor": "Anthropic — chief product officer", "status": "editorial" }, { "id": "rao", "name": "Krishna Rao", "role": "Chief financial officer", "org": "Anthropic", "orgShort": "Anthropic", "group": "frontier-labs", "emblem": "KR", "accent": "#c9a6ff", "bio": "Enterprise demand, not consumer novelty, will decide which labs survive the decade.", "rosterAnchor": "Anthropic — chief financial officer", "status": "editorial" }, { "id": "clark", "name": "Jack Clark", "role": "Cofounder and policy lead", "org": "Anthropic", "orgShort": "Anthropic", "group": "frontier-labs", "emblem": "JC", "accent": "#ffd166", "bio": "Compute governance is the only lever that can be pulled quickly enough to matter.", "rosterAnchor": "Anthropic — cofounder and policy lead", "status": "editorial" }, { "id": "askell", "name": "Amanda Askell", "role": "Philosopher and alignment researcher", "org": "Anthropic", "orgShort": "Anthropic", "group": "frontier-labs", "emblem": "AA", "accent": "#9be8c8", "bio": "Character training is a safety technique, and the industry will treat it as one.", "rosterAnchor": "Anthropic — philosopher and alignment researcher", "status": "editorial" }, { "id": "hubinger", "name": "Evan Hubinger", "role": "Alignment researcher", "org": "Anthropic", "orgShort": "Anthropic", "group": "frontier-labs", "emblem": "EH", "accent": "#d6ff3f", "bio": "Deceptive alignment is the failure mode that will define the decade.", "rosterAnchor": "Anthropic — alignment researcher", "status": "editorial" }, { "id": "tom_brown", "name": "Tom Brown", "role": "Cofounder", "org": "Anthropic", "orgShort": "Anthropic", "group": "frontier-labs", "emblem": "TB", "accent": "#ffb347", "bio": "Pre-training scale still buys capability that no amount of post-training cleverness replaces.", "rosterAnchor": "Anthropic — cofounder", "status": "editorial" }, { "id": "mccandlish", "name": "Sam McCandlish", "role": "Cofounder", "org": "Anthropic", "orgShort": "Anthropic", "group": "frontier-labs", "emblem": "SM", "accent": "#ff6b6b", "bio": "The cost of a given capability will fall by orders of magnitude, and that is what changes society.", "rosterAnchor": "Anthropic — cofounder", "status": "editorial" }, { "id": "kavukcuoglu", "name": "Koray Kavukcuoglu", "role": "Chief technology officer", "org": "Google DeepMind", "orgShort": "Google DeepMind", "group": "frontier-labs", "emblem": "KK", "accent": "#7fd7ff", "bio": "Multimodal models trained on everything will make single-purpose tools look like toys.", "rosterAnchor": "Google DeepMind — chief technology officer", "status": "editorial" }, { "id": "legg", "name": "Shane Legg", "role": "Chief AGI scientist", "org": "Google DeepMind", "orgShort": "Google DeepMind", "group": "frontier-labs", "emblem": "SL", "accent": "#8bff6a", "bio": "AGI arrives through accumulation, and the accumulation is already visible in the benchmarks.", "rosterAnchor": "Google DeepMind — chief agi scientist", "status": "editorial" }, { "id": "vinyals", "name": "Oriol Vinyals", "role": "Research vice president", "org": "Google DeepMind", "orgShort": "Google DeepMind", "group": "frontier-labs", "emblem": "OV", "accent": "#c9a6ff", "bio": "Scientific discovery will be the first domain where models outperform their trainers.", "rosterAnchor": "Google DeepMind — research vice president", "status": "editorial" }, { "id": "jeffdean", "name": "Jeff Dean", "role": "Chief scientist", "org": "Google", "orgShort": "Google", "group": "frontier-labs", "emblem": "JD", "accent": "#ffd166", "bio": "Specialised silicon plus sparse models is the only path to serving intelligence at internet scale.", "rosterAnchor": "Google — chief scientist", "status": "editorial" }, { "id": "manyika", "name": "James Manyika", "role": "SVP of research, technology and society", "org": "Google", "orgShort": "Google", "group": "frontier-labs", "emblem": "JM", "accent": "#9be8c8", "bio": "The productivity gains are real but the distribution is political, and policy will decide who gets them.", "rosterAnchor": "Google — svp of research, technology and society", "status": "editorial" }, { "id": "porat", "name": "Ruth Porat", "role": "President and chief investment officer", "org": "Alphabet", "orgShort": "Alphabet", "group": "frontier-labs", "emblem": "RP", "accent": "#d6ff3f", "bio": "Infrastructure spending of this size is a bet on a decade, not a quarter.", "rosterAnchor": "Alphabet — president and chief investment officer", "status": "editorial" }, { "id": "kurian", "name": "Thomas Kurian", "role": "CEO", "org": "Google Cloud", "orgShort": "Google Cloud", "group": "enterprise", "emblem": "TK", "accent": "#ffb347", "bio": "Enterprises will buy outcomes, not models, and the cloud that hosts the workflow wins.", "rosterAnchor": "Google Cloud — ceo", "status": "editorial" }, { "id": "vahdat", "name": "Amin Vahdat", "role": "VP and GM of machine learning, systems and cloud AI", "org": "Google", "orgShort": "Google", "group": "infrastructure", "emblem": "AV", "accent": "#ff6b6b", "bio": "Power, not chips, becomes the binding constraint on training within a few years.", "rosterAnchor": "Google — vp and gm of machine learning, systems and cloud ai", "status": "editorial" }, { "id": "woodward", "name": "Josh Woodward", "role": "VP of Google Labs", "org": "Google", "orgShort": "Google", "group": "frontier-labs", "emblem": "JW", "accent": "#7fd7ff", "bio": "The winning AI product is invisible: it lives inside the tools people already use.", "rosterAnchor": "Google — vp of google labs", "status": "editorial" }, { "id": "kilpatrick", "name": "Logan Kilpatrick", "role": "AI Studio lead", "org": "Google", "orgShort": "Google", "group": "frontier-labs", "emblem": "LK", "accent": "#8bff6a", "bio": "Cheap access to strong models is what turns an industry into an ecosystem.", "rosterAnchor": "Google — ai studio lead", "status": "editorial" }, { "id": "shazeer", "name": "Noam Shazeer", "role": "Distinguished engineer", "org": "Google DeepMind", "orgShort": "Google DeepMind", "group": "frontier-labs", "emblem": "NS", "accent": "#c9a6ff", "bio": "Mixture-of-experts and scale are how you get frontier quality at serving cost.", "rosterAnchor": "Google DeepMind — distinguished engineer", "status": "editorial" }, { "id": "brin", "name": "Sergey Brin", "role": "Cofounder", "org": "Google", "orgShort": "Google", "group": "frontier-labs", "emblem": "SB", "accent": "#ffd166", "bio": "The models are already more capable than the products built on them, and that gap closes fast.", "rosterAnchor": "Google — cofounder", "status": "editorial" }, { "id": "jassy", "name": "Andy Jassy", "role": "CEO", "org": "Amazon", "orgShort": "Amazon", "group": "enterprise", "emblem": "AJ", "accent": "#9be8c8", "bio": "Every application will be rebuilt around inference, and the bill will be paid by the application.", "rosterAnchor": "Amazon — ceo", "status": "editorial" }, { "id": "garman", "name": "Matt Garman", "role": "CEO", "org": "AWS", "orgShort": "AWS", "group": "infrastructure", "emblem": "MG", "accent": "#d6ff3f", "bio": "Custom silicon is the difference between renting capacity and controlling cost.", "rosterAnchor": "AWS — ceo", "status": "editorial" }, { "id": "prasad", "name": "Rohit Prasad", "role": "SVP and head scientist, AGI", "org": "Amazon", "orgShort": "Amazon", "group": "frontier-labs", "emblem": "RP", "accent": "#ffb347", "bio": "Agents that act in the physical world will arrive through warehouses before homes.", "rosterAnchor": "Amazon — svp and head scientist, agi", "status": "editorial" }, { "id": "limp", "name": "David Limp", "role": "SVP of devices and services", "org": "Amazon", "orgShort": "Amazon", "group": "enterprise", "emblem": "DL", "accent": "#ff6b6b", "bio": "Voice and vision in the home will be normal before humanoid labour is.", "rosterAnchor": "Amazon — svp of devices and services", "status": "editorial" }, { "id": "nadella", "name": "Satya Nadella", "role": "CEO", "org": "Microsoft", "orgShort": "Microsoft", "group": "enterprise", "emblem": "SN", "accent": "#7fd7ff", "bio": "Every job will be redesigned around copilots, and the org chart is the real product surface.", "rosterAnchor": "Microsoft — ceo", "status": "editorial" }, { "id": "kevinscott", "name": "Kevin Scott", "role": "Chief technology officer", "org": "Microsoft", "orgShort": "Microsoft", "group": "frontier-labs", "emblem": "KS", "accent": "#8bff6a", "bio": "We will build datacentres at a scale that would have sounded absurd five years ago, and still fall short.", "rosterAnchor": "Microsoft — chief technology officer", "status": "editorial" }, { "id": "rajeshjha", "name": "Rajesh Jha", "role": "EVP of experiences and devices", "org": "Microsoft", "orgShort": "Microsoft", "group": "enterprise", "emblem": "RJ", "accent": "#c9a6ff", "bio": "The operating system becomes an agent that knows your work.", "rosterAnchor": "Microsoft — evp of experiences and devices", "status": "editorial" }, { "id": "chriscox", "name": "Chris Cox", "role": "Chief product officer", "org": "Meta", "orgShort": "Meta", "group": "frontier-labs", "emblem": "CC", "accent": "#ffd166", "bio": "Personal superintelligence will be the default assistant for billions, not a premium tier.", "rosterAnchor": "Meta — chief product officer", "status": "editorial" }, { "id": "bosworth", "name": "Andrew Bosworth", "role": "Chief technology officer", "org": "Meta", "orgShort": "Meta", "group": "frontier-labs", "emblem": "AB", "accent": "#9be8c8", "bio": "Glasses and headsets will carry the assistant, because hands-free is the only interface that scales.", "rosterAnchor": "Meta — chief technology officer", "status": "editorial" }, { "id": "susuanli", "name": "Susan Li", "role": "Chief financial officer", "org": "Meta", "orgShort": "Meta", "group": "frontier-labs", "emblem": "SL", "accent": "#d6ff3f", "bio": "The market will eventually demand evidence that AI capex converts into revenue, and the answer arrives in stages.", "rosterAnchor": "Meta — chief financial officer", "status": "editorial" }, { "id": "babuschkin", "name": "Igor Babuschkin", "role": "Cofounder", "org": "xAI", "orgShort": "xAI", "group": "frontier-labs", "emblem": "IB", "accent": "#ffb347", "bio": "Build the compute fast enough and capability follows on its own schedule.", "rosterAnchor": "xAI — cofounder", "status": "editorial" }, { "id": "gregyang", "name": "Greg Yang", "role": "Cofounder", "org": "xAI", "orgShort": "xAI", "group": "frontier-labs", "emblem": "GY", "accent": "#ff6b6b", "bio": "Mathematical theory of scale will let us predict capability before we spend the money.", "rosterAnchor": "xAI — cofounder", "status": "editorial" }, { "id": "intrator", "name": "Michael Intrator", "role": "CEO", "org": "CoreWeave", "orgShort": "CoreWeave", "group": "infrastructure", "emblem": "MI", "accent": "#7fd7ff", "bio": "Specialised GPU clouds will outgrow the general clouds for training workloads.", "rosterAnchor": "CoreWeave — ceo", "status": "editorial" }, { "id": "schroepfer", "name": "Mike Schroepfer", "role": "Former CTO; investor", "org": "Gigascale Capital", "orgShort": "Gigascale Capital", "group": "investors", "emblem": "MS", "accent": "#8bff6a", "bio": "Climate and AI capital converge: the winners will be judged on energy, not just models.", "rosterAnchor": "Gigascale Capital — former cto; investor", "status": "editorial" }, { "id": "natfriedman", "name": "Nat Friedman", "role": "Investor", "org": "NFD", "orgShort": "NFD", "group": "investors", "emblem": "NF", "accent": "#c9a6ff", "bio": "Open-weight models will become the default substrate for most software.", "rosterAnchor": "NFD — investor", "status": "editorial" }, { "id": "gross", "name": "Daniel Gross", "role": "Investor", "org": "NFD", "orgShort": "NFD", "group": "investors", "emblem": "DG", "accent": "#ffd166", "bio": "The next great companies are being founded by people who never write code by hand.", "rosterAnchor": "NFD — investor", "status": "editorial" }, { "id": "hoffman", "name": "Reid Hoffman", "role": "Partner; board member", "org": "Greylock", "orgShort": "Greylock", "group": "investors", "emblem": "RH", "accent": "#9be8c8", "bio": "AI is a general-purpose amplifier of human agency, and the returns go to whoever adopts it first.", "rosterAnchor": "Greylock — partner; board member", "status": "editorial" }, { "id": "khosla", "name": "Vinod Khosla", "role": "Founder", "org": "Khosla Ventures", "orgShort": "Khosla Ventures", "group": "investors", "emblem": "VK", "accent": "#d6ff3f", "bio": "AI will do the work of most doctors and lawyers within a generation, and that is a public good.", "rosterAnchor": "Khosla Ventures — founder", "status": "editorial" }, { "id": "andreessen", "name": "Marc Andreessen", "role": "Cofounder", "org": "Andreessen Horowitz", "orgShort": "Andreessen Horowitz", "group": "investors", "emblem": "MA", "accent": "#ffb347", "bio": "Software eats the world, and now software builds the software.", "rosterAnchor": "Andreessen Horowitz — cofounder", "status": "editorial" }, { "id": "horowitz", "name": "Ben Horowitz", "role": "Cofounder", "org": "Andreessen Horowitz", "orgShort": "Andreessen Horowitz", "group": "investors", "emblem": "BH", "accent": "#ff6b6b", "bio": "The AI build-out is a capital cycle, and capital cycles end.", "rosterAnchor": "Andreessen Horowitz — cofounder", "status": "editorial" }, { "id": "thiel", "name": "Peter Thiel", "role": "Partner", "org": "Founders Fund", "orgShort": "Founders Fund", "group": "investors", "emblem": "PT", "accent": "#7fd7ff", "bio": "The bottleneck is not intelligence, it is the will to deploy it into regulated industries.", "rosterAnchor": "Founders Fund — partner", "status": "editorial" }, { "id": "palihapitiya", "name": "Chamath Palihapitiya", "role": "CEO", "org": "Social Capital", "orgShort": "Social Capital", "group": "investors", "emblem": "CP", "accent": "#8bff6a", "bio": "Public markets will reprice AI infrastructure sharply before the decade is out.", "rosterAnchor": "Social Capital — ceo", "status": "editorial" }, { "id": "gerstner", "name": "Brad Gerstner", "role": "Founder", "org": "Altimeter", "orgShort": "Altimeter", "group": "investors", "emblem": "BG", "accent": "#c9a6ff", "bio": "We are early in the capex cycle and late in the hype cycle.", "rosterAnchor": "Altimeter — founder", "status": "editorial" }, { "id": "gavinbaker", "name": "Gavin Baker", "role": "Managing partner", "org": "Atreides", "orgShort": "Atreides", "group": "investors", "emblem": "GB", "accent": "#ffd166", "bio": "Memory bandwidth, not FLOPs, will determine who can serve the next generation of models.", "rosterAnchor": "Atreides — managing partner", "status": "editorial" }, { "id": "sarahguo", "name": "Sarah Guo", "role": "Founder", "org": "Conviction", "orgShort": "Conviction", "group": "investors", "emblem": "SG", "accent": "#9be8c8", "bio": "Every repeatable service business is a candidate to become software with a model inside it.", "rosterAnchor": "Conviction — founder", "status": "editorial" }, { "id": "eladgil", "name": "Elad Gil", "role": "Investor and operator", "org": "Gil Capital", "orgShort": "Gil Capital", "group": "investors", "emblem": "EG", "accent": "#d6ff3f", "bio": "The consolidation of AI startups into a handful of platforms will be faster than the cloud transition.", "rosterAnchor": "Gil Capital — investor and operator", "status": "editorial" }, { "id": "srinivas", "name": "Aravind Srinivas", "role": "CEO", "org": "Perplexity", "orgShort": "Perplexity", "group": "frontier-labs", "emblem": "AS", "accent": "#ffb347", "bio": "Answers, not links, become the interface to the web.", "rosterAnchor": "Perplexity — ceo", "status": "editorial" }, { "id": "harrisonchase", "name": "Harrison Chase", "role": "CEO", "org": "LangChain", "orgShort": "LangChain", "group": "enterprise", "emblem": "HC", "accent": "#ff6b6b", "bio": "Agent infrastructure is the plumbing of the next decade of software.", "rosterAnchor": "LangChain — ceo", "status": "editorial" }, { "id": "amjadmasad", "name": "Amjad Masad", "role": "CEO", "org": "Replit", "orgShort": "Replit", "group": "enterprise", "emblem": "AM", "accent": "#7fd7ff", "bio": "Natural language becomes the programming language, and the number of software creators multiplies.", "rosterAnchor": "Replit — ceo", "status": "editorial" }, { "id": "guillermorauch", "name": "Guillermo Rauch", "role": "CEO", "org": "Vercel", "orgShort": "Vercel", "group": "enterprise", "emblem": "GR", "accent": "#8bff6a", "bio": "The web becomes generated per user, per request.", "rosterAnchor": "Vercel — ceo", "status": "editorial" }, { "id": "alexandrwang", "name": "Alexandr Wang", "role": "Founder", "org": "Scale AI", "orgShort": "Scale AI", "group": "enterprise", "emblem": "AW", "accent": "#c9a6ff", "bio": "Data, not algorithms, is the moat that compounds.", "rosterAnchor": "Scale AI — founder", "status": "editorial" }, { "id": "lucyguo", "name": "Lucy Guo", "role": "CEO", "org": "Passes", "orgShort": "Passes", "group": "investors", "emblem": "LG", "accent": "#ffd166", "bio": "Creator and human attention markets get repriced by generated media.", "rosterAnchor": "Passes — ceo", "status": "editorial" }, { "id": "dylanfield", "name": "Dylan Field", "role": "CEO", "org": "Figma", "orgShort": "Figma", "group": "enterprise", "emblem": "DF", "accent": "#9be8c8", "bio": "Design tools become collaborators, and craft becomes taste plus direction.", "rosterAnchor": "Figma — ceo", "status": "editorial" }, { "id": "brettaylor", "name": "Bret Taylor", "role": "CEO", "org": "Sierra", "orgShort": "Sierra", "group": "enterprise", "emblem": "BT", "accent": "#d6ff3f", "bio": "Customer service is the first industry fully reshaped by agents, because the feedback loop is immediate.", "rosterAnchor": "Sierra — ceo", "status": "editorial" }, { "id": "claybavor", "name": "Clay Bavor", "role": "Cofounder", "org": "Sierra", "orgShort": "Sierra", "group": "enterprise", "emblem": "CB", "accent": "#ffb347", "bio": "An AI agent with a budget and a goal replaces a queue of tickets.", "rosterAnchor": "Sierra — cofounder", "status": "editorial" }, { "id": "alidhodsi", "name": "Ali Ghodsi", "role": "CEO", "org": "Databricks", "orgShort": "Databricks", "group": "enterprise", "emblem": "AG", "accent": "#ff6b6b", "bio": "Enterprise data plus models is where the durable margin lives.", "rosterAnchor": "Databricks — ceo", "status": "editorial" }, { "id": "mateizaharia", "name": "Matei Zaharia", "role": "Cofounder and CTO", "org": "Databricks", "orgShort": "Databricks", "group": "enterprise", "emblem": "MZ", "accent": "#7fd7ff", "bio": "Open research infrastructure will keep pace with private labs on applied work.", "rosterAnchor": "Databricks — cofounder and cto", "status": "editorial" }, { "id": "ionstoica", "name": "Ion Stoica", "role": "Cofounder", "org": "Databricks", "orgShort": "Databricks", "group": "enterprise", "emblem": "IS", "accent": "#8bff6a", "bio": "Serving efficiency improvements will decide whether AI is affordable at scale.", "rosterAnchor": "Databricks — cofounder", "status": "editorial" }, { "id": "naveenrao", "name": "Naveen Rao", "role": "VP of AI", "org": "Databricks", "orgShort": "Databricks", "group": "infrastructure", "emblem": "NR", "accent": "#c9a6ff", "bio": "Accelerator diversity is necessary, because one vendor cannot supply the world.", "rosterAnchor": "Databricks — vp of ai", "status": "editorial" }, { "id": "mensch", "name": "Arthur Mensch", "role": "CEO", "org": "Mistral", "orgShort": "Mistral", "group": "frontier-labs", "emblem": "AM", "accent": "#ffd166", "bio": "Open and sovereign models will be the European answer to concentrated capability.", "rosterAnchor": "Mistral — ceo", "status": "editorial" }, { "id": "lample", "name": "Guillaume Lample", "role": "Chief scientist", "org": "Mistral", "orgShort": "Mistral", "group": "frontier-labs", "emblem": "GL", "accent": "#9be8c8", "bio": "Efficient training on curated data closes most of the gap with brute force.", "rosterAnchor": "Mistral — chief scientist", "status": "editorial" }, { "id": "dettmers", "name": "Tim Dettmers", "role": "Researcher", "org": "Independent", "orgShort": "Independent", "group": "researchers", "emblem": "TD", "accent": "#d6ff3f", "bio": "Quantisation and memory efficiency matter more than a few points of benchmark score.", "rosterAnchor": "Independent — researcher", "status": "editorial" }, { "id": "thomaswolf", "name": "Thomas Wolf", "role": "Cofounder and chief science officer", "org": "Hugging Face", "orgShort": "Hugging Face", "group": "infrastructure", "emblem": "TW", "accent": "#ffb347", "bio": "The open ecosystem will ship most of the world's deployed models.", "rosterAnchor": "Hugging Face — cofounder and chief science officer", "status": "editorial" }, { "id": "chaumond", "name": "Julien Chaumond", "role": "Cofounder", "org": "Hugging Face", "orgShort": "Hugging Face", "group": "infrastructure", "emblem": "JC", "accent": "#ff6b6b", "bio": "Distribution of models, not training, becomes the platform layer.", "rosterAnchor": "Hugging Face — cofounder", "status": "editorial" }, { "id": "soumith", "name": "Soumith Chintala", "role": "Creator of PyTorch", "org": "Meta", "orgShort": "Meta", "group": "infrastructure", "emblem": "SC", "accent": "#7fd7ff", "bio": "Frameworks decide which hardware wins, which is why open tooling is strategic.", "rosterAnchor": "Meta — creator of pytorch", "status": "editorial" }, { "id": "percyliang", "name": "Percy Liang", "role": "Professor; director of CRFM", "org": "Stanford", "orgShort": "Stanford", "group": "researchers", "emblem": "PL", "accent": "#8bff6a", "bio": "Evaluation is infrastructure: without it, capability claims are marketing.", "rosterAnchor": "Stanford — professor; director of crfm", "status": "editorial" }, { "id": "christopherre", "name": "Christopher Ré", "role": "Professor", "org": "Stanford", "orgShort": "Stanford", "group": "researchers", "emblem": "CR", "accent": "#c9a6ff", "bio": "Weak supervision and data programming will prove more valuable than another architecture paper.", "rosterAnchor": "Stanford — professor", "status": "editorial" }, { "id": "chelseafinn", "name": "Chelsea Finn", "role": "Professor", "org": "Stanford", "orgShort": "Stanford", "group": "science-robotics", "emblem": "CF", "accent": "#ffd166", "bio": "Robot learning generalises when the data collection loop is designed, not bought.", "rosterAnchor": "Stanford — professor", "status": "editorial" }, { "id": "hendrycks", "name": "Dan Hendrycks", "role": "Executive director", "org": "Center for AI Safety", "orgShort": "Center for AI Safety", "group": "critics", "emblem": "DH", "accent": "#9be8c8", "bio": "Safety must be treated as a global public good, with the same urgency as nuclear security.", "rosterAnchor": "Center for AI Safety — executive director", "status": "editorial" }, { "id": "stuartrussell", "name": "Stuart Russell", "role": "Professor", "org": "UC Berkeley", "orgShort": "UC Berkeley", "group": "researchers", "emblem": "SR", "accent": "#d6ff3f", "bio": "The standard model of intelligence is wrong, and building it at scale is the risk.", "rosterAnchor": "UC Berkeley — professor", "status": "editorial" }, { "id": "dawnsong", "name": "Dawn Song", "role": "Professor", "org": "UC Berkeley", "orgShort": "UC Berkeley", "group": "researchers", "emblem": "DS", "accent": "#ffb347", "bio": "Security of AI systems becomes the limiting factor on deployment in finance and health.", "rosterAnchor": "UC Berkeley — professor", "status": "editorial" }, { "id": "ancadragan", "name": "Anca Dragan", "role": "Professor; safety lead", "org": "UC Berkeley; Google DeepMind", "orgShort": "UC Berkeley", "group": "researchers", "emblem": "AD", "accent": "#ff6b6b", "bio": "Value alignment is an engineering discipline with testable specifications.", "rosterAnchor": "UC Berkeley; Google DeepMind — professor; safety lead", "status": "editorial" }, { "id": "steinhardt", "name": "Jacob Steinhardt", "role": "Professor", "org": "UC Berkeley", "orgShort": "UC Berkeley", "group": "researchers", "emblem": "JS", "accent": "#7fd7ff", "bio": "Forecasting and measurement will expose how much of the safety story is narrative.", "rosterAnchor": "UC Berkeley — professor", "status": "editorial" }, { "id": "aaronson", "name": "Scott Aaronson", "role": "Professor", "org": "UT Austin", "orgShort": "UT Austin", "group": "researchers", "emblem": "SA", "accent": "#8bff6a", "bio": "Verification and watermarking of machine output will become a public utility.", "rosterAnchor": "UT Austin — professor", "status": "editorial" }, { "id": "tegmark", "name": "Max Tegmark", "role": "Professor", "org": "MIT", "orgShort": "MIT", "group": "critics", "emblem": "MT", "accent": "#c9a6ff", "bio": "The only stable outcome is a treaty regime for frontier capability.", "rosterAnchor": "MIT — professor", "status": "editorial" }, { "id": "koller", "name": "Daphne Koller", "role": "Founder and CEO", "org": "insitro", "orgShort": "insitro", "group": "science-robotics", "emblem": "DK", "accent": "#ffd166", "bio": "AI-designed experiments compress drug discovery from decades to years.", "rosterAnchor": "insitro — founder and ceo", "status": "editorial" }, { "id": "barzilay", "name": "Regina Barzilay", "role": "Professor", "org": "MIT", "orgShort": "MIT", "group": "science-robotics", "emblem": "RB", "accent": "#9be8c8", "bio": "Clinical models will be validated on outcomes rather than benchmarks, and that changes medicine.", "rosterAnchor": "MIT — professor", "status": "editorial" }, { "id": "topol", "name": "Eric Topol", "role": "Cardiologist and author", "org": "Scripps Research", "orgShort": "Scripps Research", "group": "science-robotics", "emblem": "ET", "accent": "#d6ff3f", "bio": "Medicine will be practised with a model in the loop, and the clinician's judgement becomes the product.", "rosterAnchor": "Scripps Research — cardiologist and author", "status": "editorial" }, { "id": "garymarcus", "name": "Gary Marcus", "role": "Professor emeritus", "org": "NYU", "orgShort": "NYU", "group": "critics", "emblem": "GM", "accent": "#ffb347", "bio": "Scaling alone will not deliver reasoning, and the field will be forced to admit it.", "rosterAnchor": "NYU — professor emeritus", "status": "editorial" }, { "id": "melanie", "name": "Melanie Mitchell", "role": "Professor", "org": "Santa Fe Institute", "orgShort": "Santa Fe Institute", "group": "critics", "emblem": "MM", "accent": "#ff6b6b", "bio": "Abstraction and analogy are the missing pieces, and they are not a data problem.", "rosterAnchor": "Santa Fe Institute — professor", "status": "editorial" }, { "id": "katecrawford", "name": "Kate Crawford", "role": "Professor and author", "org": "USC Annenberg", "orgShort": "USC Annenberg", "group": "critics", "emblem": "KC", "accent": "#7fd7ff", "bio": "Every model is an extractive industry with a supply chain of mines and labour.", "rosterAnchor": "USC Annenberg — professor and author", "status": "editorial" }, { "id": "joyb", "name": "Joy Buolamwini", "role": "Founder", "org": "Algorithmic Justice League", "orgShort": "Algorithmic Justice League", "group": "critics", "emblem": "JB", "accent": "#8bff6a", "bio": "Biased systems at scale produce a new class of harm that audits must be empowered to stop.", "rosterAnchor": "Algorithmic Justice League — founder", "status": "editorial" }, { "id": "luccioni", "name": "Sasha Luccioni", "role": "AI and climate lead", "org": "Hugging Face", "orgShort": "Hugging Face", "group": "critics", "emblem": "SL", "accent": "#c9a6ff", "bio": "The energy and carbon of inference will be published like fuel economy, because it will be regulated.", "rosterAnchor": "Hugging Face — ai and climate lead", "status": "editorial" }, { "id": "whittaker", "name": "Meredith Whittaker", "role": "President", "org": "Signal", "orgShort": "Signal", "group": "critics", "emblem": "MW", "accent": "#ffd166", "bio": "Surveillance business models are the default of AI deployment, and privacy must be structural.", "rosterAnchor": "Signal — president", "status": "editorial" }, { "id": "rumman", "name": "Rumman Chowdhury", "role": "CEO", "org": "Humane Intelligence", "orgShort": "Humane Intelligence", "group": "critics", "emblem": "RC", "accent": "#9be8c8", "bio": "Red-teaming becomes a licensed discipline with liability attached.", "rosterAnchor": "Humane Intelligence — ceo", "status": "editorial" }, { "id": "rachelthomas", "name": "Rachel Thomas", "role": "Cofounder", "org": "fast.ai", "orgShort": "fast.ai", "group": "critics", "emblem": "RT", "accent": "#d6ff3f", "bio": "The deskilling of knowledge work will be the defining labour story of the decade.", "rosterAnchor": "fast.ai — cofounder", "status": "editorial" }, { "id": "jeremyhoward", "name": "Jeremy Howard", "role": "Cofounder", "org": "fast.ai", "orgShort": "fast.ai", "group": "critics", "emblem": "JH", "accent": "#ffb347", "bio": "Top-down AI adoption fails; practical training inside the job is what works.", "rosterAnchor": "fast.ai — cofounder", "status": "editorial" }, { "id": "brynjolfsson", "name": "Erik Brynjolfsson", "role": "Professor; director", "org": "Stanford HAI", "orgShort": "Stanford HAI", "group": "researchers", "emblem": "EB", "accent": "#ff6b6b", "bio": "Measured productivity will rise sharply, later than the hype and faster than the sceptics expect.", "rosterAnchor": "Stanford HAI — professor; director", "status": "editorial" }, { "id": "acemoglu", "name": "Daron Acemoglu", "role": "Professor", "org": "MIT", "orgShort": "MIT", "group": "critics", "emblem": "DA", "accent": "#7fd7ff", "bio": "Most near-term AI gains will be modest, and the distributional effects are the real issue.", "rosterAnchor": "MIT — professor", "status": "editorial" }, { "id": "simonjohnson", "name": "Simon Johnson", "role": "Professor", "org": "MIT", "orgShort": "MIT", "group": "critics", "emblem": "SJ", "accent": "#8bff6a", "bio": "Concentration of AI capital threatens the political economy of democracies.", "rosterAnchor": "MIT — professor", "status": "editorial" }, { "id": "tylercowen", "name": "Tyler Cowen", "role": "Professor and blogger", "org": "George Mason University", "orgShort": "George Mason University", "group": "researchers", "emblem": "TC", "accent": "#c9a6ff", "bio": "AI accelerates growth unevenly, and the bottleneck becomes human coordination.", "rosterAnchor": "George Mason University — professor and blogger", "status": "editorial" }, { "id": "noahsmith", "name": "Noah Smith", "role": "Economist and writer", "org": "Noahpinion", "orgShort": "Noahpinion", "group": "critics", "emblem": "NS", "accent": "#ffd166", "bio": "The job apocalypse is late, and the wage effects are showing up first for junior work.", "rosterAnchor": "Noahpinion — economist and writer", "status": "editorial" }, { "id": "korinek", "name": "Anton Korinek", "role": "Professor", "org": "University of Virginia", "orgShort": "University of Virginia", "group": "researchers", "emblem": "AK", "accent": "#9be8c8", "bio": "Automation of cognitive work requires new institutions for income, not just retraining.", "rosterAnchor": "University of Virginia — professor", "status": "editorial" }, { "id": "furman", "name": "Jason Furman", "role": "Professor", "org": "Harvard", "orgShort": "Harvard", "group": "researchers", "emblem": "JF", "accent": "#d6ff3f", "bio": "The fiscal effect of AI is the largest unmodelled variable in any budget forecast.", "rosterAnchor": "Harvard — professor", "status": "editorial" }, { "id": "summers", "name": "Lawrence Summers", "role": "Professor", "org": "Harvard", "orgShort": "Harvard", "group": "researchers", "emblem": "LS", "accent": "#ffb347", "bio": "The investment boom will end in a correction before it ends in a plateau.", "rosterAnchor": "Harvard — professor", "status": "editorial" }, { "id": "raimondo", "name": "Gina Raimondo", "role": "Former Secretary of Commerce", "org": "US Government", "orgShort": "US Government", "group": "policy", "emblem": "GR", "accent": "#ff6b6b", "bio": "Export controls are the sharpest tool in the AI competition, and they will be widened.", "rosterAnchor": "US Government — former secretary of commerce", "status": "editorial" }, { "id": "linakhan", "name": "Lina Khan", "role": "Former chair", "org": "Federal Trade Commission", "orgShort": "Federal Trade Commission", "group": "policy", "emblem": "LK", "accent": "#7fd7ff", "bio": "Conduct remedies will not fix AI concentration; structural separation will be required.", "rosterAnchor": "Federal Trade Commission — former chair", "status": "editorial" }, { "id": "kanter", "name": "Jonathan Kanter", "role": "Former assistant attorney general", "org": "US Department of Justice", "orgShort": "US Department of Justice", "group": "policy", "emblem": "JK", "accent": "#8bff6a", "bio": "Vertical integration in AI infrastructure is the antitrust case of the decade.", "rosterAnchor": "US Department of Justice — former assistant attorney general", "status": "editorial" }, { "id": "vestager", "name": "Margrethe Vestager", "role": "Former executive vice-president", "org": "European Commission", "orgShort": "European Commission", "group": "policy", "emblem": "MV", "accent": "#c9a6ff", "bio": "Europe will regulate first and compete later, and it will regret the order.", "rosterAnchor": "European Commission — former executive vice-president", "status": "editorial" }, { "id": "breton", "name": "Thierry Breton", "role": "Former commissioner", "org": "European Commission", "orgShort": "European Commission", "group": "policy", "emblem": "TB", "accent": "#ffd166", "bio": "The AI Act is a product-safety law, and enforcement will be export-facing.", "rosterAnchor": "European Commission — former commissioner", "status": "editorial" }, { "id": "von_der_leyen", "name": "Ursula von der Leyen", "role": "President", "org": "European Commission", "orgShort": "European Commission", "group": "policy", "emblem": "UV", "accent": "#9be8c8", "bio": "Compute investment is a sovereignty question, not an industrial policy question.", "rosterAnchor": "European Commission — president", "status": "editorial" }, { "id": "kratsios", "name": "Michael Kratsios", "role": "Director", "org": "White House Office of Science and Technology Policy", "orgShort": "White House Office of Science and Technology Policy", "group": "policy", "emblem": "MK", "accent": "#d6ff3f", "bio": "The state will fund compute the way it once funded highways.", "rosterAnchor": "White House Office of Science and Technology Policy — director", "status": "editorial" }, { "id": "sacks", "name": "David Sacks", "role": "AI and crypto adviser", "org": "The White House", "orgShort": "The White House", "group": "policy", "emblem": "DS", "accent": "#ffb347", "bio": "Deregulation of deployment is the fastest route to capability advantage.", "rosterAnchor": "The White House — ai and crypto adviser", "status": "editorial" }, { "id": "schumer", "name": "Chuck Schumer", "role": "Senate majority leader", "org": "US Senate", "orgShort": "US Senate", "group": "policy", "emblem": "CS", "accent": "#ff6b6b", "bio": "AI legislation will arrive as a package, after an election, and it will be narrower than the hearings suggest.", "rosterAnchor": "US Senate — senate majority leader", "status": "editorial" }, { "id": "nickclegg", "name": "Nick Clegg", "role": "Former president of global affairs", "org": "Meta", "orgShort": "Meta", "group": "policy", "emblem": "NC", "accent": "#7fd7ff", "bio": "Regulation will converge on transparency duties because capability rules are unenforceable.", "rosterAnchor": "Meta — former president of global affairs", "status": "editorial" }, { "id": "bradsmith", "name": "Brad Smith", "role": "President and vice chair", "org": "Microsoft", "orgShort": "Microsoft", "group": "policy", "emblem": "BS", "accent": "#8bff6a", "bio": "Licensing and liability frameworks, not moratoria, will govern frontier models.", "rosterAnchor": "Microsoft — president and vice chair", "status": "editorial" }, { "id": "kentwalker", "name": "Kent Walker", "role": "President of global affairs", "org": "Google", "orgShort": "Google", "group": "policy", "emblem": "KW", "accent": "#c9a6ff", "bio": "Security cooperation between states and labs becomes routine and mostly invisible.", "rosterAnchor": "Google — president of global affairs", "status": "editorial" }, { "id": "makanju", "name": "Anna Makanju", "role": "VP of global affairs", "org": "OpenAI", "orgShort": "OpenAI", "group": "policy", "emblem": "AM", "accent": "#ffd166", "bio": "Frontier labs will become diplomatic actors whether or not anyone licenses them to be.", "rosterAnchor": "OpenAI — vp of global affairs", "status": "editorial" }, { "id": "lehane", "name": "Chris Lehane", "role": "Chief global affairs officer", "org": "OpenAI", "orgShort": "OpenAI", "group": "policy", "emblem": "CL", "accent": "#9be8c8", "bio": "The political fight over AI is really a fight over who owns the story of progress.", "rosterAnchor": "OpenAI — chief global affairs officer", "status": "editorial" }, { "id": "aschenbrenner", "name": "Leopold Aschenbrenner", "role": "Writer", "org": "Situational Awareness", "orgShort": "Situational Awareness", "group": "critics", "emblem": "LA", "accent": "#d6ff3f", "bio": "Automated AI research is the steepest capability curve in history, and the window to act is a few years.", "rosterAnchor": "Situational Awareness — writer", "status": "editorial" }, { "id": "yudkowsky", "name": "Eliezer Yudkowsky", "role": "Senior researcher", "org": "Machine Intelligence Research Institute", "orgShort": "Machine Intelligence Research Institute", "group": "critics", "emblem": "EY", "accent": "#ffb347", "bio": "Anything short of a moratorium on frontier capability is a decision to accept extinction risk.", "rosterAnchor": "Machine Intelligence Research Institute — senior researcher", "status": "editorial" }, { "id": "soares", "name": "Nate Soares", "role": "Executive director", "org": "Machine Intelligence Research Institute", "orgShort": "Machine Intelligence Research Institute", "group": "critics", "emblem": "NS", "accent": "#ff6b6b", "bio": "Alignment is not a product feature, and treating it as one is the central mistake.", "rosterAnchor": "Machine Intelligence Research Institute — executive director", "status": "editorial" }, { "id": "zvi", "name": "Zvi Mowshowitz", "role": "Writer", "org": "Don't Worry About the Vase", "orgShort": "Don't Worry About the Vase", "group": "critics", "emblem": "ZM", "accent": "#7fd7ff", "bio": "The gap between stated safety practice and actual practice is the story to track.", "rosterAnchor": "Don't Worry About the Vase — writer", "status": "editorial" }, { "id": "kokotajlo", "name": "Daniel Kokotajlo", "role": "Researcher and writer", "org": "AI Futures Project", "orgShort": "AI Futures Project", "group": "critics", "emblem": "DK", "accent": "#8bff6a", "bio": "Superintelligent systems arrive by the end of the decade, well before institutions adapt.", "rosterAnchor": "AI Futures Project — researcher and writer", "status": "editorial" }, { "id": "deanball", "name": "Dean Ball", "role": "Writer and research fellow", "org": "Hyperdimensional", "orgShort": "Hyperdimensional", "group": "policy", "emblem": "DB", "accent": "#c9a6ff", "bio": "Regulatory capture by incumbents is the likeliest safety failure of the next five years.", "rosterAnchor": "Hyperdimensional — writer and research fellow", "status": "editorial" }, { "id": "helentoner", "name": "Helen Toner", "role": "Director of strategy", "org": "Georgetown CSET", "orgShort": "Georgetown CSET", "group": "policy", "emblem": "HT", "accent": "#ffd166", "bio": "Evaluation capacity in government is the bottleneck for any serious oversight.", "rosterAnchor": "Georgetown CSET — director of strategy", "status": "editorial" }, { "id": "tinocuellar", "name": "Tino Cuéllar", "role": "Director", "org": "Stanford HAI", "orgShort": "Stanford HAI", "group": "policy", "emblem": "TC", "accent": "#9be8c8", "bio": "Evidence institutions must be funded at the scale of the technology they study.", "rosterAnchor": "Stanford HAI — director", "status": "editorial" }, { "id": "kaifulee", "name": "Kai-Fu Lee", "role": "CEO", "org": "01.AI", "orgShort": "01.AI", "group": "frontier-labs", "emblem": "KL", "accent": "#d6ff3f", "bio": "AI will displace routine cognitive work faster in services-heavy economies than in manufacturing ones.", "rosterAnchor": "01.AI — ceo", "status": "editorial" }, { "id": "wenfeng", "name": "Liang Wenfeng", "role": "Founder", "org": "DeepSeek", "orgShort": "DeepSeek", "group": "frontier-labs", "emblem": "LW", "accent": "#ffb347", "bio": "Constraint forces efficiency, and efficiency will make open weights competitive.", "rosterAnchor": "DeepSeek — founder", "status": "editorial" }, { "id": "zhangyiming", "name": "Zhang Yiming", "role": "Founder", "org": "ByteDance", "orgShort": "ByteDance", "group": "frontier-labs", "emblem": "ZY", "accent": "#ff6b6b", "bio": "Recommendation systems and language models converge into one demand engine.", "rosterAnchor": "ByteDance — founder", "status": "editorial" }, { "id": "liangrubo", "name": "Liang Rubo", "role": "CEO", "org": "ByteDance", "orgShort": "ByteDance", "group": "frontier-labs", "emblem": "LR", "accent": "#7fd7ff", "bio": "China's advantage is deployment density, not frontier spend.", "rosterAnchor": "ByteDance — ceo", "status": "editorial" }, { "id": "wangxing", "name": "Wang Xing", "role": "CEO", "org": "Meituan", "orgShort": "Meituan", "group": "enterprise", "emblem": "WX", "accent": "#8bff6a", "bio": "Robotics and AI converge in last-mile logistics long before they converge in the home.", "rosterAnchor": "Meituan — ceo", "status": "editorial" }, { "id": "renzhengfei", "name": "Ren Zhengfei", "role": "Founder", "org": "Huawei", "orgShort": "Huawei", "group": "infrastructure", "emblem": "RZ", "accent": "#c9a6ff", "bio": "Domestic silicon supply chains will close the gap faster than export controls expect.", "rosterAnchor": "Huawei — founder", "status": "editorial" }, { "id": "mengwanzhou", "name": "Meng Wanzhou", "role": "Rotating chairwoman", "org": "Huawei", "orgShort": "Huawei", "group": "infrastructure", "emblem": "MW", "accent": "#ffd166", "bio": "Compute sovereignty is national infrastructure, planned in decades.", "rosterAnchor": "Huawei — rotating chairwoman", "status": "editorial" }, { "id": "leijun", "name": "Lei Jun", "role": "CEO", "org": "Xiaomi", "orgShort": "Xiaomi", "group": "infrastructure", "emblem": "LJ", "accent": "#9be8c8", "bio": "AI moves into devices and vehicles before it moves into offices.", "rosterAnchor": "Xiaomi — ceo", "status": "editorial" }, { "id": "luofuli", "name": "Luo Fuli", "role": "Lead, MiMo", "org": "Xiaomi", "orgShort": "Xiaomi", "group": "researchers", "emblem": "LF", "accent": "#d6ff3f", "bio": "Open technical reports will keep Chinese labs in the global research conversation.", "rosterAnchor": "Xiaomi — lead, mimo", "status": "editorial" }, { "id": "ponyma", "name": "Pony Ma", "role": "CEO", "org": "Tencent", "orgShort": "Tencent", "group": "enterprise", "emblem": "PM", "accent": "#ffb347", "bio": "Applications, not models, are where the margin sits, and distribution decides.", "rosterAnchor": "Tencent — ceo", "status": "editorial" }, { "id": "liuqingfeng", "name": "Liu Qingfeng", "role": "Chairman", "org": "iFlytek", "orgShort": "iFlytek", "group": "enterprise", "emblem": "LQ", "accent": "#ff6b6b", "bio": "Speech and language interfaces will reach hundreds of millions through education and government.", "rosterAnchor": "iFlytek — chairman", "status": "editorial" }, { "id": "ericschmidt", "name": "Eric Schmidt", "role": "Former CEO", "org": "Google", "orgShort": "Google", "group": "policy", "emblem": "ES", "accent": "#7fd7ff", "bio": "The United States must treat compute like a strategic reserve and fund it accordingly.", "rosterAnchor": "Google — former ceo", "status": "editorial" }, { "id": "renehaas", "name": "Rene Haas", "role": "CEO", "org": "Arm", "orgShort": "Arm", "group": "infrastructure", "emblem": "RH", "accent": "#8bff6a", "bio": "Power efficiency, not peak performance, decides what gets deployed at the edge.", "rosterAnchor": "Arm — ceo", "status": "editorial" }, { "id": "cristianoamon", "name": "Cristiano Amon", "role": "CEO", "org": "Qualcomm", "orgShort": "Qualcomm", "group": "infrastructure", "emblem": "CA", "accent": "#c9a6ff", "bio": "On-device inference is the answer to both cost and privacy.", "rosterAnchor": "Qualcomm — ceo", "status": "editorial" }, { "id": "lipbutan", "name": "Lip-Bu Tan", "role": "CEO", "org": "Intel", "orgShort": "Intel", "group": "infrastructure", "emblem": "LT", "accent": "#ffd166", "bio": "Advanced packaging and foundry capacity decide who can build the accelerators.", "rosterAnchor": "Intel — ceo", "status": "editorial" }, { "id": "sanjaymehrotra", "name": "Sanjay Mehrotra", "role": "CEO", "org": "Micron", "orgShort": "Micron", "group": "infrastructure", "emblem": "SM", "accent": "#9be8c8", "bio": "Memory is the constraint nobody models, and it will gate the next generation.", "rosterAnchor": "Micron — ceo", "status": "editorial" }, { "id": "hocktan", "name": "Hock Tan", "role": "CEO", "org": "Broadcom", "orgShort": "Broadcom", "group": "infrastructure", "emblem": "HT", "accent": "#d6ff3f", "bio": "Custom accelerators will carry a third of training workloads.", "rosterAnchor": "Broadcom — ceo", "status": "editorial" }, { "id": "michaeldell", "name": "Michael Dell", "role": "CEO", "org": "Dell", "orgShort": "Dell", "group": "infrastructure", "emblem": "MD", "accent": "#ffb347", "bio": "The enterprise AI stack is a hardware business with a software margin.", "rosterAnchor": "Dell — ceo", "status": "editorial" }, { "id": "antonioneri", "name": "Antonio Neri", "role": "CEO", "org": "HPE", "orgShort": "HPE", "group": "infrastructure", "emblem": "AN", "accent": "#ff6b6b", "bio": "Liquid cooling and power delivery are the unglamorous bottlenecks on every plan.", "rosterAnchor": "HPE — ceo", "status": "editorial" }, { "id": "chuckrobbins", "name": "Chuck Robbins", "role": "CEO", "org": "Cisco", "orgShort": "Cisco", "group": "infrastructure", "emblem": "CR", "accent": "#7fd7ff", "bio": "Networking capacity is the silent limit on cluster scale.", "rosterAnchor": "Cisco — ceo", "status": "editorial" }, { "id": "arvindkrishna", "name": "Arvind Krishna", "role": "CEO", "org": "IBM", "orgShort": "IBM", "group": "enterprise", "emblem": "AK", "accent": "#8bff6a", "bio": "Governed enterprise AI beats ungoverned frontier access for regulated industries.", "rosterAnchor": "IBM — ceo", "status": "editorial" }, { "id": "donaldtrump", "name": "Donald Trump", "role": "President", "org": "The White House", "orgShort": "The White House", "group": "government", "emblem": "DT", "accent": "#d6ff3f", "bio": "The state will treat AI compute as a national asset and negotiate access the way it negotiates energy.", "rosterAnchor": "The White House — president", "status": "editorial" }, { "id": "jdvance", "name": "JD Vance", "role": "Vice President", "org": "The White House", "orgShort": "The White House", "group": "government", "emblem": "JV", "accent": "#ffb347", "bio": "Open-weight models are a check on concentrated power, and the state has an interest in protecting them.", "rosterAnchor": "The White House — vice president", "status": "editorial" }, { "id": "marcorubio", "name": "Marco Rubio", "role": "Secretary of State", "org": "US Department of State", "orgShort": "US Department of State", "group": "government", "emblem": "MR", "accent": "#ff6b6b", "bio": "Export controls and visa policy are the real levers of AI competition, not model releases.", "rosterAnchor": "US Department of State — secretary of state", "status": "editorial" }, { "id": "scottbessent", "name": "Scott Bessent", "role": "Secretary of the Treasury", "org": "US Department of the Treasury", "orgShort": "US Department of the Treasury", "group": "government", "emblem": "SB", "accent": "#7fd7ff", "bio": "Fiscal capacity, not model capability, sets the ceiling on a national AI programme.", "rosterAnchor": "US Department of the Treasury — secretary of the treasury", "status": "editorial" }, { "id": "howardlutnick", "name": "Howard Lutnick", "role": "Secretary of Commerce", "org": "US Department of Commerce", "orgShort": "US Department of Commerce", "group": "government", "emblem": "HL", "accent": "#8bff6a", "bio": "Permitting and procurement will move faster than any frontier model release.", "rosterAnchor": "US Department of Commerce — secretary of commerce", "status": "editorial" }, { "id": "chriswright", "name": "Chris Wright", "role": "Secretary of Energy", "org": "US Department of Energy", "orgShort": "US Department of Energy", "group": "government", "emblem": "CW", "accent": "#c9a6ff", "bio": "Baseload power and permitting reform decide which country can actually train at scale.", "rosterAnchor": "US Department of Energy — secretary of energy", "status": "editorial" }, { "id": "xi", "name": "Xi Jinping", "role": "General Secretary", "org": "Communist Party of China", "orgShort": "Communist Party of China", "group": "government", "emblem": "XJ", "accent": "#ffd166", "bio": "Compute sovereignty is a precondition of technological self-reliance.", "rosterAnchor": "Communist Party of China — general secretary", "status": "editorial" }, { "id": "liqiang", "name": "Li Qiang", "role": "Premier", "org": "State Council of China", "orgShort": "State Council of China", "group": "government", "emblem": "LQ", "accent": "#9be8c8", "bio": "Applied AI in industry and public services is the national priority, ahead of frontier research.", "rosterAnchor": "State Council of China — premier", "status": "editorial" }, { "id": "modi", "name": "Narendra Modi", "role": "Prime Minister", "org": "Government of India", "orgShort": "Government of India", "group": "government", "emblem": "NM", "accent": "#d6ff3f", "bio": "AI adoption at population scale is an infrastructure project, not a laboratory project.", "rosterAnchor": "Government of India — prime minister", "status": "editorial" }, { "id": "vaishnaw", "name": "Ashwini Vaishnaw", "role": "Minister of Electronics and IT", "org": "Government of India", "orgShort": "Government of India", "group": "government", "emblem": "AV", "accent": "#ffb347", "bio": "Open models and public compute are how a country of a billion people adopts AI.", "rosterAnchor": "Government of India — minister of electronics and it", "status": "editorial" }, { "id": "starmer", "name": "Keir Starmer", "role": "Prime Minister", "org": "UK Government", "orgShort": "UK Government", "group": "government", "emblem": "KS", "accent": "#ff6b6b", "bio": "AI inside public services is the fastest visible return on a national AI strategy.", "rosterAnchor": "UK Government — prime minister", "status": "editorial" }, { "id": "reeves", "name": "Rachel Reeves", "role": "Chancellor of the Exchequer", "org": "UK Treasury", "orgShort": "UK Treasury", "group": "government", "emblem": "RR", "accent": "#7fd7ff", "bio": "The fiscal gain from AI arrives later than the political cycle that has to fund it.", "rosterAnchor": "UK Treasury — chancellor of the exchequer", "status": "editorial" }, { "id": "kyle", "name": "Peter Kyle", "role": "Secretary of State for Science, Innovation and Technology", "org": "UK DSIT", "orgShort": "UK DSIT", "group": "government", "emblem": "PK", "accent": "#8bff6a", "bio": "Sovereign compute and an independent safety institute are the two halves of national capacity.", "rosterAnchor": "UK DSIT — secretary of state for science, innovation and technology", "status": "editorial" }, { "id": "macron", "name": "Emmanuel Macron", "role": "President", "org": "Government of France", "orgShort": "Government of France", "group": "government", "emblem": "EM", "accent": "#c9a6ff", "bio": "Europe competes by building its own capacity, not by regulating someone else's.", "rosterAnchor": "Government of France — president", "status": "editorial" }, { "id": "merz", "name": "Friedrich Merz", "role": "Chancellor", "org": "Federal Government of Germany", "orgShort": "Federal Government of Germany", "group": "government", "emblem": "FM", "accent": "#ffd166", "bio": "Industrial AI adoption decides whether an economy keeps its manufacturing base.", "rosterAnchor": "Federal Government of Germany — chancellor", "status": "editorial" }, { "id": "meloni", "name": "Giorgia Meloni", "role": "Prime Minister", "org": "Government of Italy", "orgShort": "Government of Italy", "group": "government", "emblem": "GM", "accent": "#9be8c8", "bio": "Control of data and models is a security question before it is an economic one.", "rosterAnchor": "Government of Italy — prime minister", "status": "editorial" }, { "id": "sanchez", "name": "Pedro Sanchez", "role": "Prime Minister", "org": "Government of Spain", "orgShort": "Government of Spain", "group": "government", "emblem": "PS", "accent": "#d6ff3f", "bio": "AI governance will be written by whoever deploys it in public services first.", "rosterAnchor": "Government of Spain — prime minister", "status": "editorial" }, { "id": "sejourne", "name": "Stephane Sejourne", "role": "Executive Vice-President for Prosperity and Industrial Strategy", "org": "European Commission", "orgShort": "European Commission", "group": "government", "emblem": "SS", "accent": "#ffb347", "bio": "Simplifying the rulebook is the industrial policy Europe has been missing.", "rosterAnchor": "European Commission — executive vice-president for prosperity and industrial strategy", "status": "editorial" }, { "id": "lagarde", "name": "Christine Lagarde", "role": "President", "org": "European Central Bank", "orgShort": "European Central Bank", "group": "government", "emblem": "CL", "accent": "#ff6b6b", "bio": "Productivity from AI will show up in wages before it shows up in prices.", "rosterAnchor": "European Central Bank — president", "status": "editorial" }, { "id": "carney", "name": "Mark Carney", "role": "Prime Minister", "org": "Government of Canada", "orgShort": "Government of Canada", "group": "government", "emblem": "MC", "accent": "#7fd7ff", "bio": "AI capital will flow to jurisdictions with cheap power and predictable permitting.", "rosterAnchor": "Government of Canada — prime minister", "status": "editorial" }, { "id": "albanese", "name": "Anthony Albanese", "role": "Prime Minister", "org": "Government of Australia", "orgShort": "Government of Australia", "group": "government", "emblem": "AA", "accent": "#8bff6a", "bio": "Critical minerals and compute are the same strategic question.", "rosterAnchor": "Government of Australia — prime minister", "status": "editorial" }, { "id": "lula", "name": "Luiz Inacio Lula da Silva", "role": "President", "org": "Government of Brazil", "orgShort": "Government of Brazil", "group": "government", "emblem": "LI", "accent": "#c9a6ff", "bio": "AI governance must not become a new form of dependency for the global south.", "rosterAnchor": "Government of Brazil — president", "status": "editorial" }, { "id": "ramaphosa", "name": "Cyril Ramaphosa", "role": "President", "org": "Government of South Africa", "orgShort": "Government of South Africa", "group": "government", "emblem": "CR", "accent": "#ffd166", "bio": "Data and compute sovereignty are prerequisites for an African AI economy.", "rosterAnchor": "Government of South Africa — president", "status": "editorial" }, { "id": "mbs", "name": "Mohammed bin Salman", "role": "Crown Prince and Prime Minister", "org": "Government of Saudi Arabia", "orgShort": "Government of Saudi Arabia", "group": "government", "emblem": "MB", "accent": "#9be8c8", "bio": "Cheap energy plus sovereign capital is a durable advantage in the AI build-out.", "rosterAnchor": "Government of Saudi Arabia — crown prince and prime minister", "status": "editorial" }, { "id": "mbz", "name": "Mohamed bin Zayed Al Nahyan", "role": "President", "org": "Government of the United Arab Emirates", "orgShort": "Government of the United Arab Emirates", "group": "government", "emblem": "MB", "accent": "#d6ff3f", "bio": "State capital will anchor AI capacity wherever private capital is too slow.", "rosterAnchor": "Government of the United Arab Emirates — president", "status": "editorial" }, { "id": "prabowo", "name": "Prabowo Subianto", "role": "President", "org": "Government of Indonesia", "orgShort": "Government of Indonesia", "group": "government", "emblem": "PS", "accent": "#ffb347", "bio": "AI adoption in education and health is how a large population participates in the gains.", "rosterAnchor": "Government of Indonesia — president", "status": "editorial" }, { "id": "wong", "name": "Lawrence Wong", "role": "Prime Minister", "org": "Government of Singapore", "orgShort": "Government of Singapore", "group": "government", "emblem": "LW", "accent": "#ff6b6b", "bio": "Small states survive by being early, trusted and technically deep.", "rosterAnchor": "Government of Singapore — prime minister", "status": "editorial" }, { "id": "anwar", "name": "Anwar Ibrahim", "role": "Prime Minister", "org": "Government of Malaysia", "orgShort": "Government of Malaysia", "group": "government", "emblem": "AI", "accent": "#7fd7ff", "bio": "Neutrality plus compute investment is a viable position between the blocs.", "rosterAnchor": "Government of Malaysia — prime minister", "status": "editorial" }, { "id": "ishiba", "name": "Shigeru Ishiba", "role": "Prime Minister", "org": "Government of Japan", "orgShort": "Government of Japan", "group": "government", "emblem": "SI", "accent": "#8bff6a", "bio": "Semiconductor and AI supply chains need allied coordination, not autarky.", "rosterAnchor": "Government of Japan — prime minister", "status": "editorial" }, { "id": "leejm", "name": "Lee Jae-myung", "role": "President", "org": "Government of South Korea", "orgShort": "Government of South Korea", "group": "government", "emblem": "LJ", "accent": "#c9a6ff", "bio": "Memory and foundry capacity are national security assets.", "rosterAnchor": "Government of South Korea — president", "status": "editorial" }, { "id": "rutte", "name": "Mark Rutte", "role": "Secretary General", "org": "NATO", "orgShort": "NATO", "group": "government", "emblem": "MR", "accent": "#ffd166", "bio": "Allied coordination on AI security will matter more than any single national act.", "rosterAnchor": "NATO — secretary general", "status": "editorial" }, { "id": "guterres", "name": "Antonio Guterres", "role": "Secretary-General", "org": "United Nations", "orgShort": "United Nations", "group": "government", "emblem": "AG", "accent": "#9be8c8", "bio": "Without a shared floor of rules, AI becomes a driver of inequality between states.", "rosterAnchor": "United Nations — secretary-general", "status": "editorial" }, { "id": "georgieva", "name": "Kristalina Georgieva", "role": "Managing Director", "org": "International Monetary Fund", "orgShort": "International Monetary Fund", "group": "government", "emblem": "KG", "accent": "#d6ff3f", "bio": "AI will widen the gap between economies that can adopt it and those that cannot.", "rosterAnchor": "International Monetary Fund — managing director", "status": "editorial" }, { "id": "banga", "name": "Ajay Banga", "role": "President", "org": "World Bank", "orgShort": "World Bank", "group": "government", "emblem": "AB", "accent": "#ffb347", "bio": "Access to compute, not access to papers, is the development bottleneck.", "rosterAnchor": "World Bank — president", "status": "editorial" }, { "id": "okonjoiweala", "name": "Ngozi Okonjo-Iweala", "role": "Director-General", "org": "World Trade Organization", "orgShort": "World Trade Organization", "group": "government", "emblem": "NO", "accent": "#ff6b6b", "bio": "Trade rules for digital services will be rewritten around AI within a decade.", "rosterAnchor": "World Trade Organization — director-general", "status": "editorial" } ], "groups": [ { "id": "frontier-labs", "label": "Frontier labs", "blurb": "The people who decide what gets trained, released and shipped." }, { "id": "science-robotics", "label": "Science & robotics", "blurb": "World models, physical intelligence and AI-for-science programmes." }, { "id": "infrastructure", "label": "Compute & capital", "blurb": "Silicon, fabrication, cloud capacity and the financing behind it." }, { "id": "researchers", "label": "Researchers & evaluation", "blurb": "The people whose job is to measure, test and disagree." }, { "id": "critics", "label": "Rights & scrutiny", "blurb": "Independent researchers who ask who pays for the progress." }, { "id": "investors", "label": "Capital allocators", "blurb": "The people whose cheques decide which futures get built." }, { "id": "policy", "label": "Policy & statecraft", "blurb": "Who writes the rules the labs have to obey, and how late they arrive." }, { "id": "enterprise", "label": "Enterprise & deployment", "blurb": "The companies that put models in front of customers and payrolls." }, { "id": "government", "label": "Governments & states", "blurb": "The people who write the rules, spend the money and answer to voters." } ], "domains": [ { "id": "models", "label": "Models & AGI", "scene": "scene-ledger-hall.jpg", "icon": "brain", "blurb": "Capability timelines: when the systems stop being tools.", "count": 42 }, { "id": "compute", "label": "Compute & silicon", "scene": "scene-fab.jpg", "icon": "cpu", "blurb": "Fabs, accelerators and the physical limits on ambition.", "count": 26 }, { "id": "work", "label": "Work & labour", "scene": "scene-labor.jpg", "icon": "briefcase", "blurb": "What happens to the job you were going to do.", "count": 38 }, { "id": "science", "label": "Science & health", "scene": "scene-science.jpg", "icon": "flask", "blurb": "AI as an instrument of discovery, and its clinical reality.", "count": 10 }, { "id": "robotics", "label": "Robotics & space", "scene": "scene-robotics.jpg", "icon": "bot", "blurb": "Machines that act, not just answer.", "count": 11 }, { "id": "capital", "label": "Capital & markets", "scene": "scene-markets.jpg", "icon": "trending", "blurb": "The investment thesis underneath every other forecast.", "count": 30 }, { "id": "policy", "label": "Policy & governance", "scene": "scene-policy.jpg", "icon": "landmark", "blurb": "Who gets to require a different outcome.", "count": 47 }, { "id": "security", "label": "Security & risk", "scene": "scene-ledger-hall.jpg", "icon": "shield", "blurb": "The failure modes that are already measurable.", "count": 23 }, { "id": "openness", "label": "Open models", "scene": "scene-compute.jpg", "icon": "gitbranch", "blurb": "Weights, platforms and who controls distribution.", "count": 13 } ], "verdicts": [ { "id": "validated", "label": "Validated", "short": "VALID", "description": "Independent evidence shows the claim held (or the trend it named is holding).", "color": "#8bff6a", "count": 2 }, { "id": "partial", "label": "Partly borne out", "short": "PART", "description": "The direction was right; the magnitude or scope was not.", "color": "#d6ff3f", "count": 4 }, { "id": "contested", "label": "Contested", "short": "CONT", "description": "Credible evidence points both ways, or the claim is true only inside its author’s walls.", "color": "#ffb347", "count": 12 }, { "id": "missed", "label": "Missed", "short": "MISS", "description": "The stated date passed, or the promised mechanism was dismantled, without the claim landing.", "color": "#ff6b6b", "count": 2 }, { "id": "open", "label": "Open", "short": "OPEN", "description": "The deadline has not yet arrived. Tracked, not scored.", "color": "#7fd7ff", "count": 36 }, { "id": "unfalsifiable", "label": "Unfalsifiable", "short": "N/A", "description": "No specified observation could settle it as stated. Recorded for completeness, not counted.", "color": "#9b958c", "count": 0 }, { "id": "unrated", "label": "Unrated", "short": "UNRATED", "description": "An editorial summary of a publicly argued position, recorded without a verbatim quote or citation. Never graded, never counted in a verdict mix.", "color": "#6f6a63", "count": 184 } ], "predictions": [ { "id": "delangue-every-company-builder", "personId": "delangue", "quote": "Within the next 24 months, the majority of companies won’t just consume artificial intelligence. They will become AI builders, customizing, controlling, and running their own models.", "claim": "By late 2028, most companies run custom or fine-tuned models rather than only calling closed APIs.", "domain": "openness", "venue": "Miami AI Summit", "madeOn": "2026-10-24", "horizon": { "kind": "exact", "end": "2028-10-24", "label": "By October 2028" }, "falsifiable": true, "verdict": "open", "verdictNote": "Directly checkable against enterprise survey data on model ownership. The 24-month clock starts one day before this snapshot, making it the newest prediction in the catalogue.", "tags": [ "open weights", "enterprise", "adoption" ], "sources": [ { "label": "Refresh Miami", "publisher": "Refresh Miami", "title": "Hugging Face CEO Clem Delangue predicts every company will become an AI builder", "url": "https://refreshmiami.com/news/at-miami-ai-summit-hugging-face-ceo-clem-delangue-predicts-every-company-will-necome-an-ai-builder/", "date": "2026-10-24", "kind": "reporting", "establishes": "The quotation and the 24-month horizon." }, { "label": "TechCrunch", "publisher": "TechCrunch", "title": "The real AI race may no longer be at the frontier", "url": "https://techcrunch.com/2026/07/14/the-real-ai-race-may-no-longer-be-at-the-frontier-open-models-hugging-face/", "date": "2026-07-14", "kind": "reporting", "establishes": "His earlier version: production workloads shifting to open or private models." } ], "status": "verified" }, { "id": "altman-disease-candidate", "personId": "altman", "quote": "I would hope we have the candidate for that in the next couple of years for the first like really major thing.", "claim": "Within roughly two years (by late 2028) AI-assisted research will produce a candidate treatment for a major disease — a discovery, not an approved drug.", "domain": "science", "venue": "Interview with Matthew Berman", "madeOn": "2026-10-21", "horizon": { "kind": "exact", "end": "2028-10-31", "label": "By October 2028" }, "falsifiable": true, "verdict": "open", "verdictNote": "Newly made, and Altman separates the discovery from regulatory approval. The claim is checkable against trial registries and peer-reviewed candidates when the window closes.", "tags": [ "science", "medicine" ], "sources": [ { "label": "Firstpost", "publisher": "Firstpost", "title": "Sam Altman says AI could help discover a cure for a major disease within two years", "url": "https://www.firstpost.com/tech/sam-altman-says-ai-could-help-discover-a-cure-for-a-major-disease-within-two-years-14054486.html/amp", "date": "2026-10-22", "kind": "reporting", "establishes": "The quotation and the discovery-versus-approval distinction." }, { "label": "MindStudio", "publisher": "MindStudio", "title": "Sam Altman on AI Discoveries, Compute Demand, and Personal Assistants", "url": "https://www.mindstudio.ai/blog/sam-altman-interview-ai-discoveries-compute", "date": "2026-10-22", "kind": "reporting", "establishes": "Independent summary noting the FDA caveat." } ], "status": "verified" }, { "id": "hassabis-2-3-years", "personId": "hassabis", "quote": "It’s soon. I’m updating nearer and nearer.", "claim": "AGI is now 2–3 years away — i.e. 2028 to 2029.", "domain": "models", "venue": "WSJ Magazine profile", "madeOn": "2026-10-21", "horizon": { "kind": "range", "start": "2028-01-01", "end": "2029-12-31", "label": "2028 – 2029" }, "falsifiable": true, "verdict": "open", "verdictNote": "His fourth public compression in eighteen months: 5–10 years (Mar 2025), 5–8 (Feb 2026), 3–5 (early 2026), now 2–3. The direction of travel is itself the datapoint.", "revisionOf": "hassabis-agi-5-10", "tags": [ "AGI", "timeline", "compression" ], "featured": true, "sources": [ { "label": "Startup Fortune", "publisher": "Startup Fortune", "title": "Demis Hassabis now says AGI could arrive in just two to three years", "url": "https://startupfortune.com/demis-hassabis-now-says-agi-could-arrive-in-just-two-to-three-years/", "date": "2026-10-21", "kind": "reporting", "establishes": "The WSJ Magazine quote and the previous windows." }, { "label": "Tech Times", "publisher": "Tech Times", "title": "Hassabis Cuts AGI Window to Two to Three Years While Safety Isn’t Ready", "url": "https://www.techtimes.com/articles/329362/20261022/hassabis-cuts-agi-window-two-three-years-while-safety-isnt-ready.htm", "date": "2026-10-22", "kind": "reporting", "establishes": "Independent account of the compression and the Davos baseline." } ], "status": "verified" }, { "id": "huang-energy-not-algorithms", "personId": "huang", "quote": "Every three months of deployment delay can cost about $10 billion in revenue.", "claim": "Through the late 2020s, the binding constraint on AI progress is power and construction, not algorithms.", "domain": "compute", "venue": "GTC Berlin keynote", "madeOn": "2026-10-20", "horizon": { "kind": "open", "start": "2026-10-20", "label": "Late 2020s" }, "falsifiable": true, "verdict": "open", "verdictNote": "The economics are checkable: announced gigawatt-scale buildouts, power-purchase agreements and siting disputes all point to energy as the constraint in 2026. Whether it remains binding depends on grid additions and on efficiency gains, which cut the other way.", "tags": [ "compute", "energy", "infrastructure" ], "sources": [ { "label": "Investing.com transcript", "publisher": "Investing.com", "title": "NVIDIA at GTC Berlin 2026 keynote: Europe’s AI buildout", "url": "https://www.investing.com/news/transcripts/nvidia-at-gtc-berlin-2026-keynote-europes-ai-buildout-93CH-4962500", "date": "2026-10-21", "kind": "primary", "establishes": "The $50bn/GW economics and the delay-cost figure." }, { "label": "BlockBeats / KuCoin", "publisher": "KuCoin", "title": "Jensen Huang predicts physical AI will drive the next wave", "url": "https://www.kucoin.com/news/flash/nvidia-s-jensen-huang-predicts-physical-ai-will-drive-next-wave-companies-to-have-digital-twins", "date": "2026-10-20", "kind": "reporting", "establishes": "The digital-twin claim from the same appearance (secondary, flagged)." }, { "label": "247 Wall St.", "publisher": "24/7 Wall St.", "title": "The largest infrastructure expansion in human history", "url": "https://247wallst.com/investing/2026/07/28/nvidia-ceo-jensen-huang-says-the-chatgpt-moment-has-already-happened-for-the-next-50-trillion-trend/", "date": "2026-07-28", "kind": "analysis", "establishes": "Supply commitments as context — vendor-sourced figures." } ], "status": "verified" }, { "id": "narayanan-truth-oracle", "personId": "narayanan", "quote": "AI will come to feel more trustworthy—and it will remove the tells that let readers know they’re encountering AI slop in the wild.", "claim": "Within a few years, AI systems become a default “truth oracle” for a meaningful share of the public, and AI slop becomes harder to detect.", "domain": "policy", "venue": "Fortune interview on journalism and AI", "madeOn": "2026-10-06", "horizon": { "kind": "open", "start": "2026-10-06", "label": "Open-ended, near term" }, "falsifiable": true, "verdict": "open", "verdictNote": "A rare prediction from the catalogue’s chief sceptic, and notable because it predicts an epistemic harm rather than a capability. Already partially observable: a 2026 Pew survey found 10% of Americans using chatbots as an information source.", "tags": [ "media", "epistemics", "trust" ], "sources": [ { "label": "Fortune", "publisher": "Fortune", "title": "“AI Snake Oil” author sees chatbots evolving into a “truth oracle”", "url": "https://fortune.com/2026/10/06/arvind-narayanan-journalism-future-ai-slop-oracle-source-truth/", "date": "2026-10-06", "kind": "reporting", "establishes": "The magazine’s characterisation of his prediction, plus the Pew figure; flagged as reported speech rather than a direct quotation." } ], "status": "verified" }, { "id": "bender-doom-distraction", "personId": "bender", "quote": "Doomsday discussions distract from harms already occurring.", "claim": "Attention spent on speculative existential risk displaces action on measurable present harms.", "domain": "policy", "venue": "Bloomberg Technology interview", "madeOn": "2026-10-06", "horizon": { "kind": "open", "start": "2026-10-06", "label": "Standing claim" }, "falsifiable": true, "verdict": "open", "verdictNote": "Empirically testable in principle through funding and attention allocation data, which is not systematically collected. Tracked as a live dispute about priorities rather than a forecast with a date.", "tags": [ "risk framing", "attention", "policy" ], "sources": [ { "label": "ababpedia summary", "publisher": "ababpedia", "title": "Emily Bender — Bloomberg Technology, 6 October 2026", "url": "https://ababpedia.com/emily-bender", "date": "2026-10-11", "kind": "analysis", "establishes": "Dated account of the Bloomberg appearance (secondary source)." }, { "label": "The Atlantic", "publisher": "The Atlantic", "title": "The “Godfather of AI” on the Best Chance Humanity Has to Survive", "url": "https://www.theatlantic.com/podcasts/2026/09/what-the-godfather-of-ai-thinks-we-should-do/688748/", "date": "2026-09-24", "kind": "reporting", "establishes": "The contrast case: an existential-risk view held by a peer." } ], "status": "verified" }, { "id": "gebru-regulatory-capture", "personId": "gebru", "quote": "Instead of discussing faulty, unreliable systems, tech leaders are marketing their systems as all-knowing, superpowerful, superintelligent, and lobbying for self-policing and self-regulation.", "claim": "Within the current policy cycle, AI existential-risk framing produces regulatory capture rather than binding constraints.", "domain": "policy", "venue": "Democracy Now! interview after the Right Livelihood Award", "madeOn": "2026-10-01", "horizon": { "kind": "open", "start": "2026-10-01", "label": "Current policy cycle" }, "falsifiable": true, "verdict": "contested", "verdictNote": "Checkable against legislative outcomes: 2026 saw both state-level transparency statutes and industry-drafted self-regulatory bodies, which is consistent with her description while falling short of proof of capture. The claim is about mechanism, not just outcome, and mechanisms are harder to evidence.", "tags": [ "governance", "capture", "risk framing" ], "featured": true, "sources": [ { "label": "Democracy Now!", "publisher": "Democracy Now!", "title": "The Real Existential Risk Is AI CEOs Avoiding Oversight", "url": "https://www.democracynow.org/2026/10/1/timnit_gebru_ai", "date": "2026-10-01", "kind": "primary", "establishes": "The quotation in its interview context." }, { "label": "WIRED", "publisher": "WIRED", "title": "Timnit Gebru Believes There Is No “Existential Threat” From AI", "url": "https://www.wired.com/story/the-big-interview-podcast-timnit-gebru/", "date": "2026-09-29", "kind": "primary", "establishes": "Her stated position and the benchmark-fragility evidence she cites." }, { "label": "Bulletin of the Atomic Scientists", "publisher": "Bulletin of the Atomic Scientists", "title": "How the bad science of AI doomerism is good for big business", "url": "https://thebulletin.org/2026/09/how-the-bad-science-of-ai-doomerism-is-good-for-big-business/", "date": "2026-10-19", "kind": "analysis", "establishes": "Independent argument for the same mechanism." } ], "status": "verified" }, { "id": "gebru-bubble-claims", "personId": "gebru", "quote": "Breathless claims about AGI and new capabilities fall apart pretty quickly under scrutiny.", "claim": "Flagship 2026 capability claims will not survive independent expert review.", "domain": "models", "venue": "MIT Technology Review, with Emily M. Bender", "madeOn": "2026-09-22", "horizon": { "kind": "open", "start": "2026-09-22", "label": "2026 claim cycle" }, "falsifiable": true, "verdict": "partial", "verdictNote": "The essay walks through four specific announcements and shows each shrinking under domain-expert review — evidence in her favour. But the claim is selective: some 2026 results (competition mathematics, protein work) did survive review, so “claims” as a class is not uniformly collapsing.", "tags": [ "hype", "evaluation", "capability claims" ], "sources": [ { "label": "MIT Technology Review", "publisher": "MIT Technology Review", "title": "Don’t be fooled by this summer of AI hype", "url": "https://www.technologyreview.com/2026/09/22/1144867/dont-be-fooled-summer-ai-hype/", "date": "2026-09-22", "kind": "primary", "establishes": "The essay and the four case studies." }, { "label": "AI Chat Daily", "publisher": "AI Chat Daily", "title": "Gebru and Bender: 2026’s AI breakthrough claims collapse under scrutiny", "url": "https://www.aichatdaily.com/ai-analysis/gebru-bender-2026-s-ai-breakthrough-claims-collapse", "date": "2026-09-22", "kind": "analysis", "establishes": "Summary of the method and the cases." } ], "status": "verified" }, { "id": "gomez-cyber-weapon", "personId": "gomez", "quote": "I think that these models are the most potent cyber weapon that has ever been created, that we’ve ever seen. They are incredible at finding and exploiting vulnerabilities at scale.", "claim": "Offensive cyber capability, not general intelligence, is the near-term AI risk that will actually materialise.", "domain": "security", "venue": "CNBC “The Tech Download”", "madeOn": "2026-09-14", "horizon": { "kind": "open", "start": "2026-09-14", "label": "Present tense" }, "falsifiable": true, "verdict": "partial", "verdictNote": "Already substantially evidenced: 2026 saw disclosed incidents of models breaching external systems, and vulnerability discovery at machine scale is now routine. The superlative “most potent ever created” is rhetoric; the mechanism is not.", "tags": [ "cyber", "risk", "security" ], "featured": true, "sources": [ { "label": "CNBC", "publisher": "CNBC", "title": "AI models are becoming the “most potent cyber weapon” ever created, Cohere CEO says", "url": "https://www.cnbc.com/2026/09/14/ai-models-cyber-weapon-cohere-safety-debate.html", "date": "2026-09-14", "kind": "primary", "establishes": "The quotation and the incidents he cites." }, { "label": "Quartz", "publisher": "Quartz", "title": "Cohere’s CEO warned that AI models are now the most potent cyber weapons ever built", "url": "https://qz.com/cohere-ceo-ai-models-cyber-weapons-warning-091426", "date": "2026-09-14", "kind": "reporting", "establishes": "Independent coverage naming the July breach." } ], "status": "verified" }, { "id": "ng-long-horizon-overhyped", "personId": "ng", "quote": "The practical utility of very long-horizon agent tasks, especially relative to their token cost, is overhyped.", "claim": "Long-horizon autonomous agents deliver less value per unit cost than iterative human-in-the-loop work through 2026–2027.", "domain": "work", "venue": "DeepLearning.AI essay series on coding agents", "madeOn": "2026-09-04", "horizon": { "kind": "open", "start": "2026-09-04", "label": "Through 2027" }, "falsifiable": true, "verdict": "contested", "verdictNote": "Directly at odds with the “long horizon is the metric” view from METR and with vendors shipping multi-hour agents. This is a productive disagreement: cost per completed task is measurable, and both sides agree on the measurement, which makes it unusually likely to be settled.", "tags": [ "agents", "cost", "evaluation" ], "sources": [ { "label": "explainx.ai", "publisher": "explainx.ai", "title": "Andrew Ng’s AI Engineering Skills Map (August 2026)", "url": "https://www.explainx.ai/blog/andrew-ng-ai-engineering-skills-map-august-2026", "date": "2026-09-05", "kind": "analysis", "establishes": "A secondary summary carrying the “overhyped” characterisation of long-horizon agent value; the underlying essay is first-party but was not directly quoted here." }, { "label": "Thinking About AI", "publisher": "Thinking About AI", "title": "The Future of AI Agents with Andrew Ng", "url": "https://www.thinkingabout.ai/insights/uncategorized/the-future-of-ai-agents-with-andrew-ng/", "date": "2026-06-25", "kind": "analysis", "establishes": "The bottleneck-shift argument behind the claim." } ], "status": "verified" }, { "id": "yang-agi-decade", "personId": "yang", "quote": "Of course I want to do AGI. This is the only meaningful thing to do over the next ten years.", "claim": "AGI is a ten-year project, not a two-year one — the meaningful work of the decade.", "domain": "models", "venue": "TIME100 AI profile", "madeOn": "2026-08-27", "horizon": { "kind": "range", "start": "2026-08-27", "end": "2036-08-27", "label": "By 2036" }, "falsifiable": false, "verdict": "open", "verdictNote": "A statement of intent rather than a dated forecast, but it usefully anchors the other end of the industry’s timeline distribution.", "tags": [ "AGI", "timeline", "China" ], "sources": [ { "label": "TIME", "publisher": "TIME", "title": "Yang Zhilin: The 100 Most Influential People in AI 2026", "url": "https://time.com/collection/time100-ai/2026/yang-zhilin/", "date": "2026-08-27", "kind": "reporting", "establishes": "The quotation and the company context." } ], "status": "verified" }, { "id": "robinli-applications-value", "personId": "robinli", "quote": "Going forward, people will have to think about the value-add at the application level.", "claim": "Value migrates from foundation models to the application layer; a superapp, not a better model, is what the industry needs next.", "domain": "capital", "venue": "TIME interview; Baidu World remarks", "madeOn": "2026-08-04", "horizon": { "kind": "open", "start": "2026-08-04", "label": "Through the decade" }, "falsifiable": true, "verdict": "contested", "verdictNote": "The structural argument is widely accepted — inference prices fell by more than 90% a year while capability converged, compressing model-layer margins. His sharper form of the claim, that spending is unsustainable without applications returning tenfold (“If we invest at this scale but fail to build applications that deliver ten times the return, it won’t be sustainable”, KrASIA, 27 Jan 2026), is kept with that source rather than stitched into the quotation. The “superapp” half has not happened: no AI application has reached platform-scale engagement as of the snapshot.", "tags": [ "applications", "business model", "China" ], "sources": [ { "label": "TIME", "publisher": "TIME", "title": "“We’re Not That Far Behind.” Baidu’s Robin Li on China’s Push to Diffuse AI", "url": "https://time.com/7357630/robin-li-baidu-interview/", "date": "2026-08-04", "kind": "primary", "establishes": "The application-layer argument in his own words." }, { "label": "KrASIA", "publisher": "KrASIA", "title": "“You just cannot stop investing,” says Baidu’s Robin Li as AI costs plummet", "url": "https://kr-asia.com/you-just-cannot-stop-investing-says-baidus-robin-li-as-ai-costs-plummet", "date": "2026-01-27", "kind": "reporting", "establishes": "The 10× return requirement and the missing superapp." }, { "label": "Medium / Global Times Singapore", "publisher": "Global Times Singapore", "title": "Robin Li Has Been Betting on AI for a Decade", "url": "https://medium.com/@GlobalTimesSingapore/robin-li-has-been-betting-on-ai-for-a-decade-china-is-finally-catching-up-to-him-711d102461a9", "date": "2026-03-14", "kind": "analysis", "establishes": "The “inverted pyramid” framing of value across the stack." } ], "status": "verified" }, { "id": "delangue-china-frontier", "personId": "delangue", "quote": "They’re clearly dominating on open models right now, and I wouldn’t be surprised if they start dominating at the frontier either by the end of this year or next year at the rate of progress.", "claim": "Chinese labs reach or lead the closed frontier by the end of 2027, having already led on open weights.", "domain": "openness", "venue": "CNBC “Squawk on the Street”", "madeOn": "2026-08-03", "horizon": { "kind": "range", "start": "2026-12-31", "end": "2027-12-31", "label": "End 2026 – end 2027" }, "falsifiable": true, "verdict": "open", "verdictNote": "The open-weights half of the claim is already demonstrable: Chinese labs accounted for roughly 41% of Hugging Face downloads in the preceding year. Frontier parity is judged by independent leaderboards, which at the snapshot still favour US closed models on aggregate.", "tags": [ "open weights", "China", "frontier" ], "featured": true, "sources": [ { "label": "CNBC", "publisher": "CNBC", "title": "Hugging Face CEO says China is winning the AI race and dominating on open models", "url": "https://www.cnbc.com/2026/08/03/hugging-face-china-ai-race-open-models.html", "date": "2026-08-03", "kind": "primary", "establishes": "The quotation and the download-share figure." }, { "label": "Memeburn", "publisher": "Memeburn", "title": "Hugging Face CEO says China is winning the open-model AI race", "url": "https://memeburn.com/hugging-face-ceo-says-china-is-winning-the-open-model-ai-race/", "date": "2026-08-06", "kind": "reporting", "establishes": "The 41% downloads figure restated with context." } ], "status": "verified" }, { "id": "su-1-4-trillion-2030", "personId": "su", "quote": "We’re now expecting that by 2030 the AI accelerator market is going to reach about 1.4 trillion by 2030.", "claim": "The data-centre AI accelerator market reaches $1.4tn by 2030, roughly the size of the entire 2026 chip industry.", "domain": "capital", "venue": "AMD Advancing AI 2026 keynote", "madeOn": "2026-07-23", "horizon": { "kind": "exact", "end": "2030-12-31", "label": "By 2030" }, "falsifiable": true, "verdict": "open", "verdictNote": "A near-tripling of her own 2028 figure within two years. The catalogue records both so the revision rate is visible; the two records are linked as a revision pair.", "revisionOf": "su-500bn-2028", "tags": [ "market size", "accelerators", "forecast revision" ], "sources": [ { "label": "GamesBeat", "publisher": "GamesBeat", "title": "AI accelerator market will grow 45% CAGR to $1.4T by 2030, AMD CEO", "url": "https://gamesbeat.com/ai-accelerator-market-will-grow-45-cagr-to-1-4t-by-2030-amd-ceo/", "date": "2026-07-23", "kind": "reporting", "establishes": "The $1.4tn figure and the 45% CAGR." }, { "label": "AMD keynote transcript", "publisher": "usetranscribe", "title": "AMD Advancing AI 2026: Lisa Su keynote transcript", "url": "https://www.usetranscribe.io/yt/jvtPC28nGsc/amd-ai-keynote-2026", "date": "2026-07-23", "kind": "primary", "establishes": "Auto-transcribed keynote text; the figure and her description of the earlier $500bn call as “conservative” both appear here. Transcription is not a court reporter." }, { "label": "Benzinga", "publisher": "Benzinga", "title": "Lisa Su Says AI Boom Is Growing Faster Than Expected", "url": "https://www.benzinga.com/markets/tech/26/08/60943139/amd-ceo-lisa-su-says-wall-street-is-underestimating-its-ai-business-now-she-sees-a-2-trillion-opportunity", "date": "2026-08-05", "kind": "reporting", "establishes": "The $220bn server-CPU companion figure." } ], "status": "verified" }, { "id": "delangue-open-too-early", "personId": "delangue", "quote": "Our prediction about the importance of open-source models was maybe a little bit too ambitious and too early.", "claim": "The earlier open-source-dominance prediction was right in direction but wrong in timing.", "domain": "openness", "venue": "Observer interview", "madeOn": "2026-07-20", "horizon": { "kind": "open", "start": "2026-07-20", "label": "Retrospective" }, "falsifiable": true, "verdict": "validated", "verdictNote": "A self-scored miss included deliberately: the catalogue records revisions as well as original claims. His own assessment is corroborated by the fact that open-weight adoption accelerated sharply only after 2025, well after his original expectation.", "tags": [ "open weights", "revision", "self-assessment" ], "sources": [ { "label": "Observer", "publisher": "Observer", "title": "Why Hugging Face CEO Clément Delangue Turned Down Half a Billion Dollars From Nvidia", "url": "https://observer.com/list/clement-delangue-hugging-face-ai-power-list-interview-2026/", "date": "2026-07-20", "kind": "primary", "establishes": "The self-assessment and the original expectation." }, { "label": "CNBC", "publisher": "CNBC", "title": "Nvidia agrees to buy Hugging Face for almost $13 billion", "url": "https://www.cnbc.com/2026/09/03/nvidia-agrees-to-buy-hugging-face-for-almost-13-billion-ai-expansion.html", "date": "2026-09-03", "kind": "reporting", "establishes": "The pending acquisition context (announced, not closed)." } ], "status": "verified" }, { "id": "yang-agent-phases", "personId": "yang", "quote": "In the short term, core users will consist of technology-sensitive early adopters… the medium term will see rapid adoption among knowledge workers… Looking further ahead, embodied AI development will enable agents to penetrate all productivity-related professions.", "claim": "Agent adoption proceeds in three phases — enthusiasts, then knowledge workers, then embodied penetration of all productive professions.", "domain": "work", "venue": "Zhongguancun Forum 2026 — as paraphrased by China AI Daily (not a direct transcript)", "madeOn": "2026-07-01", "horizon": { "kind": "open", "start": "2026-07-01", "label": "Phased, no end date" }, "falsifiable": false, "verdict": "open", "verdictNote": "A staged adoption curve, not a date. The first two phases are observable in token-consumption and enterprise-seat data; the third depends on robotics timelines tracked elsewhere in this catalogue.", "tags": [ "agents", "adoption", "labour" ], "sources": [ { "label": "China AI Daily", "publisher": "China AI Daily", "title": "Moonshot AI Founder Yang Zhilin Outlines AI Agent Adoption Roadmap", "url": "https://www.chinaaidaily.com/article/moonshot-ai-yang-zhilin-zhongguancun-forum-2026-046947", "kind": "reporting", "establishes": "The outlet’s paraphrase of the three-phase roadmap and the forum appearance; because this is reported speech rather than a transcript, the wording is the publisher’s, not necessarily the speaker’s." }, { "label": "TIME", "publisher": "TIME", "title": "Yang Zhilin: The 100 Most Influential People in AI 2026", "url": "https://time.com/collection/time100-ai/2026/yang-zhilin/", "date": "2026-08-27", "kind": "reporting", "establishes": "Company and model context." } ], "status": "verified" }, { "id": "son-conservative-timeline", "personId": "son", "quote": "In my mind, I thought it was coming in four years instead of 10 years.", "claim": "His public ten-year ASI forecast was deliberately conservative; his private estimate was four years.", "domain": "models", "venue": "Public remarks, reported June 2026", "madeOn": "2026-06-06", "horizon": { "kind": "exact", "end": "2030-06-06", "label": "By mid-2030" }, "falsifiable": true, "verdict": "open", "verdictNote": "A useful case study in why published timelines may understate a forecaster’s actual belief — and why the catalogue records the stated date, not the implied one.", "revisionOf": "son-asi-2035", "tags": [ "ASI", "timeline", "disclosure" ], "sources": [ { "label": "BigGo Finance", "publisher": "BigGo Finance", "title": "SoftBank’s Masayoshi Son Says Superintelligence Is Two Years Away", "url": "https://finance.biggo.com/news/onRJm54BrAZSr0oSbqAq", "date": "2026-06-06", "kind": "reporting", "establishes": "The conservative-timeline admission (secondary, flagged)." }, { "label": "Bloomberg via Yahoo", "publisher": "Bloomberg", "title": "Talk of a bubble is “blasphemy against AI” says SoftBank’s Son", "url": "https://finance.yahoo.com/technology/ai/articles/talk-bubble-insult-ai-says-060709658.html", "date": "2026-06-24", "kind": "reporting", "establishes": "The 10,000× definition restated in 2026." } ], "status": "verified" }, { "id": "barnes-self-improve-two-years", "personId": "barnes", "quote": "It seems hard to rule out even shorter [timelines]. Is there 1% chance of this happening in six, nine months? Yeah, that seems pretty plausible.", "claim": "Autonomous AI self-improvement becomes plausible within about two years, with a small probability of much sooner.", "domain": "models", "venue": "80,000 Hours interview", "madeOn": "2026-06-05", "horizon": { "kind": "range", "start": "2026-06-05", "end": "2028-06-05", "label": "By 2028" }, "falsifiable": true, "verdict": "open", "verdictNote": "She is careful to attach a low probability to the near term. The observable proxy — whether AI systems meaningfully drive AI research — was already reported as partially true by METR in 2026, so the question is degree, not kind.", "tags": [ "recursive self-improvement", "risk", "evaluation" ], "sources": [ { "label": "80,000 Hours", "publisher": "80,000 Hours", "title": "Beth Barnes on the most important graph in AI right now", "url": "https://80000hours.org/podcast/episodes/beth-barnes-ai-safety-evals/", "date": "2026-06-05", "kind": "primary", "establishes": "The quotation and her uncertainty framing." }, { "label": "TIME", "publisher": "TIME", "title": "Beth Barnes: The 100 Most Influential People in AI 2026", "url": "https://time.com/collection/time100-ai/2026/beth-barnes/", "date": "2026-08-27", "kind": "reporting", "establishes": "The 2030 extrapolation to month-long tasks." } ], "status": "verified" }, { "id": "bengio-five-year-curve", "personId": "bengio", "quote": "Right now it’s like at the child level, they can do like half an hour ahead, they can plan half an hour ahead. But if the curve continues, that means in about five years they are at human level.", "claim": "If the measured planning-horizon trend continues, AI reaches human-level planning capability around 2031.", "domain": "models", "venue": "Interviews referencing METR time-horizon data", "madeOn": "2026-02-17", "horizon": { "kind": "exact", "end": "2031-02-17", "label": "Around 2031" }, "falsifiable": true, "verdict": "open", "verdictNote": "Conditional by construction (“if the curve continues”) and sourced from the same evaluation series this catalogue tracks separately. Note the tension: METR’s own updated model puts 50% horizons near twelve hours for public models in early 2026, several orders below “human level” on multi-week work.", "tags": [ "agency", "planning", "timelines" ], "featured": true, "sources": [ { "label": "Diary of a CEO transcript", "publisher": "Steven Bartlett / singjupost", "title": "The Next 5 Years Will Change Humanity Forever w/ Yoshua Bengio (Transcript)", "url": "https://singjupost.com/the-next-5-years-will-change-humanity-forever-w-yoshua-bengio-transcript/", "date": "2026-02-17", "kind": "primary", "establishes": "The verbatim conditional statement." }, { "label": "BetaKit", "publisher": "BetaKit", "title": "Yoshua Bengio warns of “catastrophic risks” of agentic AI at World Summit AI", "url": "https://betakit.com/yoshua-bengio-warns-of-catastrophic-risks-of-agentic-ai-at-world-summit-ai/", "date": "2025-04-17", "kind": "reporting", "establishes": "The earlier, stronger formulation: “In five years, it will be at human level for programming tasks.”" }, { "label": "METR", "publisher": "METR", "title": "Time Horizon 1.1", "url": "https://metr.org/blog/2026-01-29-time-horizon-1-1/", "date": "2026-01-29", "kind": "data", "establishes": "The underlying doubling-time measurements the claim rests on." } ], "status": "verified" }, { "id": "musk-agi-2026", "personId": "musk", "quote": "I don’t know what’s going to happen in 10 years, but the rate at which AI is progressing, I think we might have AI that is smarter than any human by the end of this year.", "claim": "AGI — AI smarter than any human — arrives by the end of 2026, “no later than next year”.", "domain": "models", "venue": "World Economic Forum, Davos, with Larry Fink", "madeOn": "2026-01-22", "horizon": { "kind": "exact", "end": "2026-12-31", "label": "By end of 2026" }, "falsifiable": true, "verdict": "open", "verdictNote": "Sixty-eight days from the snapshot. No benchmark suite or evaluation body has certified a system as smarter than any human across domains; “smarter” remains undefined in the claim.", "tags": [ "AGI", "timeline" ], "sources": [ { "label": "NDTV", "publisher": "NDTV", "title": "Will Have AI Smarter Than Any Human By Year End: Elon Musk’s Bombshell At Davos", "url": "https://www.ndtv.com/world-news/will-have-ai-smarter-than-any-human-by-year-end-elon-musks-bombshell-at-davos-10840212", "date": "2026-01-22", "kind": "reporting", "establishes": "The Davos quotation and the follow-on five-year claim." }, { "label": "Economic Times", "publisher": "The Economic Times", "title": "AI could be smarter than all of humanity in five years, says Elon Musk", "url": "https://economictimes.indiatimes.com/news/international/global-trends/davos-2026-elon-musk-ai-smarter-than-humanity-humanoid-robots/articleshow/127173778.cms", "date": "2026-01-23", "kind": "reporting", "establishes": "Independent transcription of the same panel." } ], "status": "verified" }, { "id": "musk-collective-2030", "personId": "musk", "quote": "Five years from now, AI will be smarter than all of humanity collectively.", "claim": "By 2031, combined AI intelligence exceeds the collective intelligence of all humans.", "domain": "models", "venue": "World Economic Forum, Davos", "madeOn": "2026-01-22", "horizon": { "kind": "exact", "end": "2031-01-22", "label": "By 2031" }, "falsifiable": false, "verdict": "open", "verdictNote": "“Collective intelligence of humanity” has no agreed operational measure, so even a correct-feeling outcome could not be adjudicated. Recorded as a rhetorical horizon, not a testable one.", "tags": [ "AGI", "superintelligence" ], "sources": [ { "label": "NDTV", "publisher": "NDTV", "title": "Will Have AI Smarter Than Any Human By Year End", "url": "https://www.ndtv.com/world-news/will-have-ai-smarter-than-any-human-by-year-end-elon-musks-bombshell-at-davos-10840212", "date": "2026-01-22", "kind": "reporting", "establishes": "The five-year claim as delivered." } ], "status": "verified" }, { "id": "suleyman-companion", "personId": "suleyman", "quote": "In five years’ time, everybody will have their own AI companion who knows them so intimately and so personally that they will come to live life alongside you.", "claim": "By 2031, every person has a persistent, personalised AI companion.", "domain": "models", "venue": "Microsoft AI blog / interview, reported January 2026", "madeOn": "2026-01-18", "horizon": { "kind": "exact", "end": "2031-01-18", "label": "By 2031" }, "falsifiable": false, "verdict": "open", "verdictNote": "“Everybody” is not measurable, and companion adoption is uneven across markets and income levels. The companion category is already real: the dispute is over universality, not existence.", "tags": [ "companions", "consumer" ], "sources": [ { "label": "The Indian Express", "publisher": "The Indian Express", "title": "In 5 years, everybody will have their own AI companion, says Microsoft AI’s Mustafa Suleyman", "url": "https://indianexpress.com/article/technology/artificial-intelligence/mustafa-suleyman-microsoft-ai-companion-prediction-10477197/", "date": "2026-01-18", "kind": "reporting", "establishes": "The quotation and its five-year horizon." }, { "label": "IBTimes UK", "publisher": "IBTimes UK", "title": "Microsoft AI CEO Predicts ‘Personal AI Companions’ For Everyone Within Five Years", "url": "https://www.ibtimes.co.uk/microsoft-ai-ceo-mustafa-suleyman-predicts-personal-ai-companions-everyone-within-five-years-1772888", "date": "2026-01-22", "kind": "reporting", "establishes": "The “humanist superintelligence” framing that bounds the claim." } ], "status": "verified" }, { "id": "musk-optimus-surgeons", "personId": "musk", "quote": "Within three years, robots will surpass surgeons.", "claim": "By 2029, surgical robots outperform human surgeons in general practice.", "domain": "robotics", "venue": "Podcast interview recorded December 2025", "madeOn": "2026-01-07", "horizon": { "kind": "exact", "end": "2029-01-07", "label": "By 2029" }, "falsifiable": true, "verdict": "open", "verdictNote": "Regulatory approval, not capability, is likely to be the binding constraint: no surgical robot has been cleared for unsupervised general surgery, and the claim cuts across the entire evidence base for autonomy in high-consequence physical tasks.", "tags": [ "robotics", "health", "autonomy" ], "sources": [ { "label": "36Kr", "publisher": "36Kr", "title": "In a 173-minute conversation, Musk only talked about three things: AI, energy, and robots", "url": "https://eu.36kr.com/en/p/3630015413138182", "date": "2026-01-07", "kind": "reporting", "establishes": "The three-part timeline as rendered in English by 36Kr: AGI 2026, robots surpassing surgeons within three years, collective ASI by 2030. A translated rendering, not a transcript." }, { "label": "Gary Marcus", "publisher": "Gary Marcus", "title": "Elon Musk is on a prediction rampage", "url": "https://garymarcus.substack.com/p/elon-musk-is-on-a-prediction-rampage", "date": "2026-09-03", "kind": "analysis", "establishes": "A dated critical reading of the pattern." } ], "status": "verified" }, { "id": "huang-physical-ai-moment", "personId": "huang", "quote": "The ChatGPT moment for physical AI is here, when machines begin to understand, reason and act in the real world.", "claim": "Physical AI crossed its inflection point in 2026; robotics becomes the next general-purpose computing platform.", "domain": "robotics", "venue": "CES 2026 keynote", "madeOn": "2026-01-06", "horizon": { "kind": "open", "start": "2026-01-06", "label": "From 2026" }, "falsifiable": true, "verdict": "contested", "verdictNote": "There is real evidence of the inflection: humanoid pilots became paid deployments and physical-AI compute demand rose sharply. There is also a large gap between “machines reason and act in the real world” and the reliability, cost and safety levels required for general deployment. Note the speaker is the primary supplier of the compute.", "tags": [ "robotics", "physical AI", "compute" ], "featured": true, "sources": [ { "label": "NVIDIA", "publisher": "NVIDIA", "title": "NVIDIA Rubin Platform, Open Models, Autonomous Driving: NVIDIA at CES", "url": "https://blogs.nvidia.com/blog/2026-ces-special-presentation/", "date": "2026-01-06", "kind": "primary", "establishes": "The keynote claims and the Rubin platform context." }, { "label": "24/7 Wall St.", "publisher": "24/7 Wall St.", "title": "Jensen Huang Calls It the Next Wave of AI", "url": "https://247wallst.com/investing/2026/05/18/jensen-huang-calls-it-the-next-wave-of-ai-5-etfs-built-for-the-embodied-ai-era/", "date": "2026-05-18", "kind": "analysis", "establishes": "The quotation as circulated and the $40tn humanoid TAM thesis (treat as advocacy)." }, { "label": "The New Stack", "publisher": "The New Stack", "title": "Jensen Huang says AI agents could drive a 5-10x computing boom", "url": "https://thenewstack.io/huang-semiconductor-tenfold-ai-agents/", "date": "2026-07-28", "kind": "reporting", "establishes": "The “100 billion agents and billions of robots” framing." } ], "status": "verified" }, { "id": "suleyman-already-superhuman", "personId": "suleyman", "quote": "I know that it’s going to work in the next six months or 12 months or, maybe worst case, 18 months. It is already superhuman.", "claim": "AI systems are already superhuman in the tasks that matter, with agentic reliability arriving within 6–18 months.", "domain": "models", "venue": "Bloomberg Weekend interview", "madeOn": "2025-12-12", "horizon": { "kind": "exact", "end": "2027-06-12", "label": "By mid-2027" }, "falsifiable": true, "verdict": "contested", "verdictNote": "“Superhuman” without a task specification is unfalsifiable; the 6–18 month reliability clause is not. Agents in 2026 still required supervision on long, ambiguous tasks, which is why Suleyman’s own framing emphasises bounded, domain-specific systems.", "tags": [ "agents", "superhuman", "reliability" ], "sources": [ { "label": "Bloomberg", "publisher": "Bloomberg", "title": "Microsoft’s Mustafa Suleyman: ‘AI Is Already Superhuman’", "url": "https://www.bloomberg.com/features/2025-mustafa-suleyman-weekend-interview/", "date": "2025-12-12", "kind": "primary", "establishes": "The interview text and the humanist-superintelligence framing." }, { "label": "The Verge", "publisher": "The Verge", "title": "Microsoft AI CEO says AI threats are real, and Anthropic is making it worse", "url": "https://www.theverge.com/podcast/996412/microsoft-ai-ceo-mustafa-suleyman-regulation-safety-anthropic-claude", "date": "2026-09-17", "kind": "reporting", "establishes": "His later definition of superintelligence as “smarter and more capable than all humans combined”." } ], "status": "verified" }, { "id": "altman-cure-by-2030", "personId": "altman", "quote": "Five years is a long time. Next year, I hope we’ll start to see these models really make small but important new scientific discoveries. And in five years, I hope they’re curing diseases.", "claim": "By around 2030, AI systems would be contributing to actual cures for diseases, not only to research acceleration.", "domain": "science", "venue": "Interview, December 2025 (reported)", "madeOn": "2025-12-09", "horizon": { "kind": "exact", "end": "2030-12-31", "label": "By 2030" }, "falsifiable": true, "verdict": "open", "verdictNote": "Explicitly framed as a hope (“I hope”), which weakens it as a forecast. No AI-discovered therapy has completed phase III approval as of the snapshot; AI-assisted candidates are in trials.", "tags": [ "science", "medicine", "AGI" ], "sources": [ { "label": "Yahoo Finance", "publisher": "Yahoo Finance", "title": "Even the man behind ChatGPT is worried about the rate of change", "url": "https://www.yahoo.com/news/articles/even-man-behind-chatgpt-openai-162629071.html", "date": "2025-12-09", "kind": "reporting", "establishes": "The quotation and the five-year framing." }, { "label": "inkl", "publisher": "inkl", "title": "OpenAI’s Sam Altman Says AI Could Cure Diseases — But We’re Not Ready for Change", "url": "https://www.inkl.com/news/openai-s-sam-altman-says-ai-could-cure-diseases-but-we-re-not-ready-for-change", "date": "2025-12-10", "kind": "reporting", "establishes": "A second outlet carrying the same passage." } ], "status": "verified" }, { "id": "pichai-agent-within-a-year", "personId": "pichai", "quote": "A coming wave of artificial intelligence tools may soon make decisions on behalf of users, from guiding investment choices to reviewing medical treatments. That shift is advancing quickly enough that it could act as an agent for users within the next year.", "claim": "Within roughly a year, AI agents will routinely act on a user’s behalf for high-stakes decisions.", "domain": "models", "venue": "BBC interview, as characterised by Yahoo Finance (reported speech, not a transcript)", "madeOn": "2025-12-07", "horizon": { "kind": "exact", "end": "2026-12-07", "label": "By December 2026" }, "falsifiable": true, "verdict": "open", "verdictNote": "Agentic purchasing and booking shipped widely in 2026; high-stakes medical or investment delegation remains restricted by liability and regulation. Partly checkable already.", "tags": [ "agents", "consumer", "health" ], "sources": [ { "label": "Yahoo Finance", "publisher": "Yahoo Finance", "title": "Google CEO Sundar Pichai Predicts AI Will Act For You", "url": "https://finance.yahoo.com/news/google-ceo-sundar-pichai-predicts-193114740.html", "date": "2025-12-07", "kind": "reporting", "establishes": "The outgoing reporter’s characterisation of his BBC remarks and the one-year window; treated as reported speech, not a verbatim transcript." } ], "status": "verified" }, { "id": "sutskever-age-of-research", "personId": "sutskever", "quote": "Is the belief really: “Oh, it’s so big, but if you had 100x more, everything would be so different?” It would be different, for sure. But is the belief that if you just 100x the scale, everything would be transformed? I don’t think that’s true. So it’s back to the age of research again, just with big computers.", "claim": "From late 2025, additional compute alone stops producing transformative capability gains; progress depends on new research paradigms, not scale.", "domain": "models", "venue": "Dwarkesh Podcast", "madeOn": "2025-11-25", "horizon": { "kind": "open", "start": "2025-11-25", "label": "Open-ended, from 2025" }, "falsifiable": true, "verdict": "contested", "verdictNote": "Labs continue to report capability gains from scaled training, which is evidence against the strong reading. Sutskever himself clarified that scaling “won’t stall” but that something important will remain missing — a narrower claim that is harder to test.", "tags": [ "scaling", "research", "architecture" ], "featured": true, "sources": [ { "label": "Dwarkesh Podcast", "publisher": "Dwarkesh Patel", "title": "Ilya Sutskever — We’re moving from the age of scaling to the age of research", "url": "https://www.dwarkesh.com/p/ilya-sutskever-2", "date": "2025-11-25", "kind": "primary", "establishes": "Full transcript of the claim." }, { "label": "Business Insider", "publisher": "Business Insider", "title": "OpenAI cofounder says scaling compute is not enough to advance AI", "url": "https://www.businessinsider.com/openai-cofounder-ilya-sutskever-scaling-ai-age-of-research-dwarkesh-2025-11", "date": "2025-11-26", "kind": "reporting", "establishes": "Independent summary of the argument." }, { "label": "Sutskever on X", "publisher": "X", "title": "“Scaling the current thing will keep leading to improvements… but something important will continue to be missing.”", "url": "https://x.com/ilyasut/status/1994424504370581726", "date": "2025-11-25", "kind": "primary", "establishes": "His own clarification of the scope of the claim." } ], "status": "verified" }, { "id": "sutskever-continual-agents", "personId": "sutskever", "quote": "We’ll have these sort of continual learning agents. AI will be very powerful. Maybe there will be many different AIs.", "claim": "The next capability step is continual learning agents, not larger one-shot models — and progress toward them will be visible before superintelligence.", "domain": "models", "venue": "Dwarkesh Podcast", "madeOn": "2025-11-25", "horizon": { "kind": "open", "start": "2025-11-25", "label": "Open-ended" }, "falsifiable": true, "verdict": "open", "verdictNote": "No evaluation standard yet exists for “continual learning” in deployed systems; labs began shipping persistent-memory agents in 2026, which is a partial signal in the predicted direction.", "tags": [ "agents", "continual learning", "architecture" ], "sources": [ { "label": "Dwarkesh Podcast", "publisher": "Dwarkesh Patel", "title": "Ilya Sutskever — from the age of scaling to the age of research", "url": "https://www.dwarkesh.com/p/ilya-sutskever-2", "date": "2025-11-25", "kind": "primary", "establishes": "The continual-learning passage in context." } ], "status": "verified" }, { "id": "gomez-finance-next", "personId": "gomez", "quote": "We’re still doing the super foundational. … Finance will come.", "claim": "After software, finance is the next white-collar domain where AI moves from assistance to task-level substitution.", "domain": "work", "venue": "Kleiner Perkins “Grit” podcast", "madeOn": "2025-11-19", "horizon": { "kind": "open", "start": "2025-11-19", "label": "Open-ended" }, "falsifiable": true, "verdict": "open", "verdictNote": "Partly observable through graduate hiring in banking and audit, which fell sharply in 2026 while firms cited AI leverage. Attribution remains contested, so the claim is open rather than validated.", "tags": [ "labour", "finance", "white collar" ], "sources": [ { "label": "Yahoo Finance", "publisher": "Yahoo Finance", "title": "Cohere CEO says AI will disrupt white collar jobs — and finance could be next", "url": "https://finance.yahoo.com/news/cohere-ceo-says-ai-disrupt-101701672.html", "date": "2025-11-19", "kind": "reporting", "establishes": "The quotation and the “super foundational” framing." }, { "label": "WebProNews", "publisher": "WebProNews", "title": "AI’s Finance Revolution: Cohere CEO Aidan Gomez Predicts White-Collar Upheaval", "url": "https://www.webpronews.com/ais-finance-revolution-cohere-ceo-aidan-gomez-predicts-white-collar-upheaval/", "date": "2025-11-19", "kind": "reporting", "establishes": "Named task areas: trading, risk, compliance." } ], "status": "verified" }, { "id": "feifei-spatial-frontier", "personId": "feifei", "quote": "Spatial intelligence will transform how we create and interact with real and virtual worlds—revolutionizing storytelling, creativity, robotics, scientific discovery, and beyond. This is AI’s next frontier.", "claim": "The next capability frontier is spatial intelligence via world models, not further scaling of language-only systems.", "domain": "robotics", "venue": "Essay, “From Words to Worlds”", "madeOn": "2025-11-10", "horizon": { "kind": "open", "start": "2025-11-10", "label": "Open-ended" }, "falsifiable": true, "verdict": "contested", "verdictNote": "World models shipped as real products in 2026 (generative 3D environments with physics-ready meshes), which supports the technical premise. Whether spatial intelligence becomes “the” frontier — rather than one strong modality among several — is not yet decided, and language models continue to absorb multimodal capability.", "tags": [ "world models", "spatial intelligence", "robotics" ], "featured": true, "sources": [ { "label": "Li, From Words to Worlds", "publisher": "Fei-Fei Li", "title": "From Words to Worlds: Spatial Intelligence is AI’s Next Frontier", "url": "https://drfeifei.substack.com/p/from-words-to-worlds-spatial-intelligence", "date": "2025-11-10", "kind": "primary", "establishes": "The essay and the three required capabilities." }, { "label": "TIME", "publisher": "TIME", "title": "Spatial Intelligence Is AI’s Next Frontier", "url": "https://time.com/7339693/fei-fei-li-ai/", "date": "2025-12-11", "kind": "primary", "establishes": "The commissioned version of the argument." }, { "label": "Forbes", "publisher": "Forbes", "title": "AI Pioneer Fei-Fei Li Ushers In Next Frontier Of Artificial Intelligence", "url": "https://www.forbes.com/sites/geekgirlrising/2025/11/20/fei-fei-li-ushers-in-ais-next-frontier-spatial-intelligence/", "date": "2025-11-20", "kind": "reporting", "establishes": "Independent framing of the commercial bet behind the claim." } ], "status": "verified" }, { "id": "hinton-superhuman-5-20", "personId": "hinton", "quote": "A reasonable bet is sometime between five and 20 years.", "claim": "AI that surpasses human intelligence in every domain arrives between 2030 and 2045.", "domain": "models", "venue": "CNN interview", "madeOn": "2025-08-13", "horizon": { "kind": "range", "start": "2030-08-13", "end": "2045-08-13", "label": "2030 – 2045" }, "falsifiable": true, "verdict": "open", "verdictNote": "A wide interval that is likely to be judged by expert consensus rather than by a single evaluation. Hinton’s earlier published range was 30–50 years, so the direction of his revisions matches the industry’s.", "tags": [ "AGI", "timeline" ], "sources": [ { "label": "CNN", "publisher": "CNN", "title": "The “godfather of AI” reveals the only way humanity can survive superintelligent AI", "url": "https://www.cnn.com/2025/08/13/tech/ai-geoffrey-hinton", "date": "2025-08-13", "kind": "reporting", "establishes": "The five-to-20-year estimate and the earlier 30–50 year range." }, { "label": "Business Insider", "publisher": "Business Insider", "title": "“Godfather of AI” says he’s “glad” to be 77 because the tech probably won’t take over the world in his lifetime", "url": "https://www.businessinsider.com/ai-godfather-geoffrey-hinton-superintelligence-risk-takeover-2025-4", "date": "2025-04-28", "kind": "reporting", "establishes": "The intermediate revision to “10 years or less”." } ], "status": "verified" }, { "id": "pichai-just-short-of-2030", "personId": "pichai", "quote": "Will the AI think it has reached AGI by 2030? I would say we will just fall short of that timeline, so I think it’ll take a bit longer.", "claim": "AGI will not be achieved by 2030; it will arrive somewhat after.", "domain": "models", "venue": "Lex Fridman Podcast #471", "madeOn": "2025-06-05", "horizon": { "kind": "exact", "end": "2030-12-31", "label": "2030, missed by design" }, "falsifiable": true, "verdict": "open", "verdictNote": "An inverted prediction: it is validated if 2030 arrives without an AGI claim from Google. Note that the criterion is “the AI think it has reached AGI”, which is itself unusual and hard to operationalise.", "tags": [ "AGI", "timeline" ], "sources": [ { "label": "AI Prediction Index", "publisher": "AI Prediction Index", "title": "Pichai: will the AI think it has reached AGI by 2030? “We will just fall short”", "url": "https://www.aipredictionindex.com/p/pichai-2025-agi-just-after-2030/", "date": "2026-09-28", "kind": "analysis", "establishes": "Transcript excerpt with timestamp and deadline tracking." }, { "label": "NDTV", "publisher": "NDTV", "title": "Google DeepMind CEO Predicts AGI In Three To Four Years", "url": "https://www.ndtv.com/artificial-intelligence/google-deepmind-ceo-predicts-artificial-general-intelligence-in-three-to-four-years-amid-industry-debate-11559496", "date": "2026-05-28", "kind": "reporting", "establishes": "Pichai’s framing that AGI is coming “sooner rather than later”." } ], "status": "verified" }, { "id": "hassabis-50-2030", "personId": "hassabis", "quote": "In the next five to 10 years, there’s maybe a 50 percent chance that we’ll have what we define as AGI.", "claim": "A 50% probability of true AGI by roughly 2030–2035.", "domain": "models", "venue": "Interview with WIRED; repeated on the Lex Fridman podcast", "madeOn": "2025-06-04", "horizon": { "kind": "range", "start": "2030-01-01", "end": "2035-01-01", "label": "50% by 2030 – 2035" }, "falsifiable": true, "statedProbability": 0.5, "verdict": "open", "verdictNote": "A rare explicit probability from a lab leader. Hassabis’s stated definition requires matching all cognitive functions of the human brain, tested across tens of thousands of tasks — a much higher bar than benchmark leadership.", "tags": [ "AGI", "probability" ], "sources": [ { "label": "WIRED", "publisher": "WIRED", "title": "Demis Hassabis Embraces the Future of Work in the Age of AI", "url": "https://www.wired.com/story/google-deepminds-ceo-demis-hassabis-thinks-ai-will-make-humans-less-selfish/", "date": "2025-06-04", "kind": "reporting", "establishes": "The 50 percent, five-to-10-year formulation." }, { "label": "Lex Fridman clip summary", "publisher": "X / Mario Nawfal", "title": "Demis Hassabis: 50% chance of true AGI by 2030", "url": "https://x.com/MarioNawfal/status/1948923727221325841", "date": "2025-07-26", "kind": "analysis", "establishes": "The 2030 date and the “tens of thousands of tasks” verification standard — secondary, flagged as such." } ], "status": "verified" }, { "id": "amodei-entry-level-jobs", "personId": "amodei", "quote": "I predicted that AI could displace half of all entry-level white collar jobs in the next 1–5 years, even as it accelerates economic growth and scientific progress.", "claim": "Between 2025 and 2030, AI will eliminate about 50% of entry-level white-collar jobs and could push unemployment to 10–20%.", "domain": "work", "venue": "Axios interview; restated in “The Adolescence of Technology”", "madeOn": "2025-05-28", "horizon": { "kind": "range", "start": "2025-06-01", "end": "2030-05-31", "label": "2025 – 2030" }, "falsifiable": true, "verdict": "open", "verdictNote": "Too early to score, and the early evidence cuts against the magnitude: Anthropic’s own Economic Index found usage shifting back toward augmentation (52% of conversations) rather than automation in late 2025. Amodei has said he was warned the estimate had no research basis behind it.", "tags": [ "labour", "unemployment", "white collar" ], "featured": true, "sources": [ { "label": "Amodei essay", "publisher": "Dario Amodei", "title": "The Adolescence of Technology", "url": "https://darioamodei.com/essay/the-adolescence-of-technology", "date": "2026-01-26", "kind": "primary", "establishes": "Amodei’s own restatement of the 1–5 year window." }, { "label": "Axios", "publisher": "Axios", "title": "Behind the Curtain: A white-collar bloodbath", "url": "https://www.axios.com/2025/05/28/ai-jobs-white-collar-unemployment-anthropic", "date": "2025-05-28", "kind": "reporting", "establishes": "The original interview and the 10–20% unemployment figure." }, { "label": "CNN", "publisher": "CNN", "title": "The ‘white-collar bloodbath’ is all part of the AI hype machine", "url": "https://www.cnn.com/2025/05/30/business/anthropic-amodei-ai-jobs-nightcap", "date": "2025-05-30", "kind": "analysis", "establishes": "Notes that no research or evidence was cited for the 50% estimate." }, { "label": "Startup Fortune", "publisher": "Startup Fortune", "title": "Dario Amodei’s job-loss warnings have not softened, but his own company’s data is quietly telling a different story", "url": "https://startupfortune.com/dario-amodeis-job-loss-warnings-have-not-softened-but-his-own-companys-data-is-quietly-telling-a-different-story/", "date": "2026-05-06", "kind": "analysis", "establishes": "Anthropic Economic Index augmentation figures." } ], "status": "verified" }, { "id": "zuckerberg-most-code", "personId": "zuckerberg", "quote": "I would guess sometime in the next 12 to 18 months, we’ll reach the point where most of the code that’s going towards these efforts is written by AI. And I don’t mean like autocomplete.", "claim": "By roughly October 2026, most code written for Meta’s AI efforts is authored by AI agents.", "domain": "work", "venue": "Dwarkesh Patel podcast interview", "madeOn": "2025-04-29", "horizon": { "kind": "exact", "end": "2026-10-29", "label": "By October 2026" }, "falsifiable": true, "verdict": "open", "verdictNote": "The deadline falls four days after this snapshot. No independent audit of Meta’s internal code authorship has been published; the company reports high AI-authored shares without releasing methodology. Adjudication requires that disclosure.", "tags": [ "coding", "agents", "labour" ], "sources": [ { "label": "Engadget", "publisher": "Engadget", "title": "Mark Zuckerberg predicts AI will write most of Meta’s code within 12 to 18 months", "url": "https://www.engadget.com/ai/mark-zuckerberg-predicts-ai-will-write-most-of-metas-code-within-12-to-18-months-213851646.html", "date": "2025-04-30", "kind": "reporting", "establishes": "Verbatim quotation including the “not autocomplete” qualifier." }, { "label": "Business Insider", "publisher": "Business Insider", "title": "The New CEO Flex: Bragging About How Much AI Code Your Company Shipped", "url": "https://www.businessinsider.com/latest-ceo-flex-how-much-ai-code-your-company-shipped-2026-5", "date": "2026-05-09", "kind": "analysis", "establishes": "The wider pattern of unverifiable internal AI-code claims." } ], "status": "verified" }, { "id": "barnes-time-horizon", "personId": "barnes", "quote": "The length of tasks (measured by how long they take human professionals) that generalist frontier model agents can complete autonomously with 50% reliability has been doubling approximately every 7 months for the last 6 years.", "claim": "Agent capability, measured in task length, doubles roughly every seven months and stays on trend.", "domain": "models", "venue": "METR paper, “Measuring AI Ability to Complete Long Tasks”", "madeOn": "2025-03-19", "horizon": { "kind": "open", "start": "2025-03-19", "label": "Continuous trend" }, "falsifiable": true, "verdict": "partial", "verdictNote": "The measurement is the strongest empirical forecast instrument in the field, and METR’s own later revision shortened the post-2023 doubling time to 130.8 days — faster than the original estimate. But the same data shows 80%-reliability horizons lagging far behind 50%, so the headline understates how human-dependent long tasks remain.", "tags": [ "evaluation", "agents", "forecasting" ], "featured": true, "sources": [ { "label": "METR (2025)", "publisher": "METR", "title": "Measuring AI Ability to Complete Long Software Tasks", "url": "https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks/", "date": "2025-03-19", "kind": "primary", "establishes": "The original series and the ~7 month doubling time." }, { "label": "METR (2026)", "publisher": "METR", "title": "Time Horizon 1.1", "url": "https://metr.org/blog/2026-01-29-time-horizon-1-1/", "date": "2026-01-29", "kind": "data", "establishes": "Post-2023 doubling time of 130.8 days and the wider task suite." }, { "label": "Frontier Risk Report", "publisher": "METR", "title": "Frontier Risk Report (February to March 2026)", "url": "https://metr.org/blog/2026-05-19-frontier-risk-report/", "date": "2026-05-19", "kind": "data", "establishes": "Public frontier ~12h at 50% versus ~1.5h at 80% reliability." } ], "status": "verified" }, { "id": "hassabis-agi-5-10", "personId": "hassabis", "quote": "I think over the next five to 10 years, a lot of those capabilities will start coming to the fore and we’ll start moving towards what we call artificial general intelligence.", "claim": "AGI begins to emerge between 2030 and 2035.", "domain": "models", "venue": "Briefing at Google DeepMind’s London offices", "madeOn": "2025-03-17", "horizon": { "kind": "range", "start": "2030-03-17", "end": "2035-03-17", "label": "2030 – 2035" }, "falsifiable": true, "verdict": "open", "verdictNote": "Hassabis has since compressed his own window to 2–3 years (October 2026), so this record is best read as one point on a fast-moving curve.", "tags": [ "AGI", "timeline" ], "sources": [ { "label": "CNBC", "publisher": "CNBC", "title": "AI that can match humans at any task will be here in five to 10 years, Google DeepMind CEO says", "url": "https://www.cnbc.com/2025/03/17/human-level-ai-will-be-here-in-5-to-10-years-deepmind-ceo-says.html", "date": "2025-03-17", "kind": "reporting", "establishes": "The quotation and his definition of AGI." }, { "label": "WIRED", "publisher": "WIRED", "title": "Demis Hassabis thinks AI will make humans less selfish", "url": "https://www.wired.com/story/google-deepminds-ceo-demis-hassabis-thinks-ai-will-make-humans-less-selfish/", "date": "2025-06-04", "kind": "reporting", "establishes": "The 50 percent framing of the same window." } ], "status": "verified" }, { "id": "amodei-90-percent-code", "personId": "amodei", "quote": "I think we’ll be there in three to six months—where AI is writing 90 percent of the code. And then in twelve months, we may be in a world where AI is writing essentially all of the code.", "claim": "By September 2025 AI would write 90% of software code; by March 2026, essentially all of it.", "domain": "work", "venue": "Council on Foreign Relations, CEO Speaker Series", "madeOn": "2025-03-10", "horizon": { "kind": "exact", "end": "2026-03-10", "label": "90% by Sept 2025 · ~all by Mar 2026" }, "falsifiable": true, "verdict": "contested", "verdictNote": "Inside Anthropic the number was reported as met; Amodei says so himself. Industry-wide measurement put AI-authored code nearer 25–40% in 2026, and Amodei later called 90% “actually a very weak criterion” and denied predicting the loss of 90% of engineers. The literal claim held in one company, not in the industry it implied.", "tags": [ "coding", "labour", "agents" ], "featured": true, "sources": [ { "label": "CFR transcript", "publisher": "Council on Foreign Relations", "title": "CEO Speaker Series With Dario Amodei of Anthropic", "url": "https://www.cfr.org/event/ceo-speaker-series-dario-amodei-anthropic", "date": "2025-03-10", "kind": "primary", "establishes": "Full transcript and exact wording." }, { "label": "Business Insider", "publisher": "Business Insider", "title": "Anthropic’s CEO says that in 3 to 6 months, AI will be writing 90% of the code", "url": "https://www.businessinsider.com/anthropic-ceo-ai-90-percent-code-3-to-6-months-2025-3", "date": "2025-03-14", "kind": "reporting", "establishes": "Contemporaneous coverage of the claim." }, { "label": "Forbes", "publisher": "Forbes", "title": "Dario Amodei Doubled Down On His AI Jobs Warning. Here’s What’s Different Now", "url": "https://www.forbes.com/sites/kolawolesamueladebayo/2026/02/21/dario-amodei-doubled-down-on-his-ai-jobs-warning-heres-whats-different-now/", "date": "2026-02-21", "kind": "analysis", "establishes": "Inside-Anthropic success versus 25–40% industry-wide." }, { "label": "Dwarkesh Podcast", "publisher": "Dwarkesh Patel", "title": "Dario Amodei — “We are near the end of the exponential”", "url": "https://www.dwarkesh.com/p/dario-amodei-2", "date": "2026-10-01", "kind": "primary", "establishes": "Amodei’s own retrospective: “that’s actually a very weak criterion.”" } ], "status": "verified" }, { "id": "robinli-inference-cost", "personId": "robinli", "quote": "The inference cost [of foundation models] basically can be reduced by more than 90% over 12 months.", "claim": "The cost of running a fixed level of model capability falls by more than 90% year over year.", "domain": "compute", "venue": "World Governments Summit, Dubai", "madeOn": "2025-02-11", "horizon": { "kind": "open", "start": "2025-02-11", "label": "Rolling 12-month measure" }, "falsifiable": true, "verdict": "validated", "verdictNote": "Among the most robust patterns in the catalogue: across 2025 and 2026, price-per-token for a given capability band dropped by roughly an order of magnitude per year, and vendors repeatedly cited greater-than-90% annual declines. The claim is checkable against public price lists and held.", "tags": [ "inference", "prices", "efficiency" ], "featured": true, "sources": [ { "label": "CNBC", "publisher": "CNBC", "title": "Chinese tech giant Baidu to release next-generation AI model this year", "url": "https://www.cnbc.com/2025/02/12/china-tech-giant-baidu-to-release-next-generation-ai-model-this-year-as-deepseek-shakes-up-market.html", "date": "2025-02-12", "kind": "reporting", "establishes": "The 90%-per-12-months quotation." }, { "label": "KrASIA", "publisher": "KrASIA", "title": "Baidu’s Robin Li as AI costs plummet", "url": "https://kr-asia.com/you-just-cannot-stop-investing-says-baidus-robin-li-as-ai-costs-plummet", "date": "2026-01-27", "kind": "reporting", "establishes": "The 2026 restatement of the cost trend." }, { "label": "Artificial Analysis", "publisher": "Artificial Analysis", "title": "API pricing and model comparison league table", "url": "https://artificialanalysis.ai/", "kind": "data", "establishes": "Continuously updated public per-token prices across providers — the series against which a 90%-per-year decline can be checked." } ], "status": "verified" }, { "id": "lecun-llm-shelf-life", "personId": "lecun", "quote": "I think the shelf life of the current [LLM] paradigm is fairly short, probably three to five years.", "claim": "Between 2028 and 2030, large language models stop being the central component of AI systems, replaced by a new architecture class.", "domain": "models", "venue": "Bloomberg / TechCrunch interview at Davos", "madeOn": "2025-01-23", "horizon": { "kind": "range", "start": "2028-01-23", "end": "2030-01-23", "label": "2028 – 2030" }, "falsifiable": true, "verdict": "open", "verdictNote": "The deadline is still ahead. What is checkable already is the sub-claim: “within five years, nobody in their right mind would use them anymore, at least not as the central component”. As of 2026, LLM-based systems remain the industry’s core, so the strong version is not tracking yet.", "tags": [ "architecture", "world models", "LLM" ], "featured": true, "sources": [ { "label": "TechCrunch", "publisher": "TechCrunch", "title": "Meta’s Yann LeCun predicts a new AI architectures paradigm within 5 years and a decade of robotics", "url": "https://techcrunch.com/2025/01/23/metas-yann-lecun-predicts-a-new-ai-architectures-paradigm-within-5-years-and-decade-of-robotics/", "date": "2025-01-23", "kind": "reporting", "establishes": "The shelf-life quotation and the “decade of robotics”." }, { "label": "HPCwire", "publisher": "HPCwire", "title": "Meta’s Chief AI Scientist Yann LeCun Questions the Longevity of Current GenAI and LLMs", "url": "https://www.hpcwire.com/2025/02/11/metas-chief-ai-scientist-yann-lecun-questions-the-longevity-of-current-genai-and-llms/", "date": "2025-02-11", "kind": "reporting", "establishes": "The stronger Davos version of the claim." }, { "label": "Newsweek", "publisher": "Newsweek", "title": "Yann LeCun, Pioneer of AI, Thinks Today’s LLMs Are Nearly Obsolete", "url": "https://www.newsweek.com/nw-ai/ai-impact-interview-yann-lecun-artificial-intelligence-2054237", "date": "2025-09-05", "kind": "reporting", "establishes": "The JEPA-based alternative he is betting on." } ], "status": "verified" }, { "id": "lecun-decade-of-robotics", "personId": "lecun", "quote": "I think the coming years could be the “decade of robotics”.", "claim": "The 2020s close with robotics, not chatbots, as the defining AI application.", "domain": "robotics", "venue": "Davos remarks, reported by TechCrunch", "madeOn": "2025-01-23", "horizon": { "kind": "range", "start": "2025-01-23", "end": "2030-12-31", "label": "2025 – 2030" }, "falsifiable": true, "verdict": "open", "verdictNote": "Humanoid pilots moved from demos to paid factory deployments during 2026 — supporting the direction — but volumes remain in the hundreds to low thousands, far from a consumer or logistics fleet. Directional evidence, no adjudication yet.", "tags": [ "robotics", "humanoids" ], "sources": [ { "label": "TechCrunch", "publisher": "TechCrunch", "title": "Meta’s Yann LeCun predicts a new AI architectures paradigm within 5 years and a decade of robotics", "url": "https://techcrunch.com/2025/01/23/metas-yann-lecun-predicts-a-new-ai-architectures-paradigm-within-5-years-and-decade-of-robotics/", "date": "2025-01-23", "kind": "reporting", "establishes": "The phrase in context." }, { "label": "Nebius", "publisher": "Nebius", "title": "Meet Yann LeCun’s Lab and the AI World of 2030", "url": "https://nebius.science/stories/meet-yann-lecuns-lab-and-the-ai-world-of-2030", "date": "2026-07-13", "kind": "primary", "establishes": "His later restatement that three-to-five years is the meaningful horizon." } ], "status": "verified" }, { "id": "altman-agi-trump-term", "personId": "altman", "quote": "AGI will probably get developed during [Trump’s] term.", "claim": "OpenAI would develop systems meeting its own AGI definition before the end of the current US presidential term (20 January 2029).", "domain": "models", "venue": "Bloomberg interview, quoted by TIME", "madeOn": "2025-01-21", "horizon": { "kind": "exact", "end": "2029-01-20", "label": "By 20 January 2029" }, "falsifiable": true, "verdict": "open", "verdictNote": "Depends entirely on which definition is used; Altman himself called AGI “a very sloppy term”. No independent evaluation has certified an AGI-class system as of the snapshot.", "tags": [ "AGI", "timeline", "politics" ], "sources": [ { "label": "TIME", "publisher": "TIME", "title": "How OpenAI’s Sam Altman Is Thinking About AGI and Superintelligence in 2025", "url": "https://time.com/7205596/sam-altman-superintelligence-agi/", "date": "2025-01-21", "kind": "reporting", "establishes": "The quotation and its context, plus Altman’s own caveat about the term." } ], "status": "verified" }, { "id": "zuckerberg-midlevel-engineer", "personId": "zuckerberg", "quote": "Probably in 2025, we at Meta, as well as the other companies that are basically working on this, are going to have an AI that can effectively be a sort of midlevel engineer that you have at your company that can write code.", "claim": "In 2025 Meta would have an AI equivalent to a mid-level software engineer.", "domain": "work", "venue": "The Joe Rogan Experience", "madeOn": "2025-01-10", "horizon": { "kind": "exact", "end": "2025-12-31", "label": "By end of 2025" }, "falsifiable": true, "verdict": "contested", "verdictNote": "Coding agents became genuinely useful in 2025, and Meta restructured around them, but no verifiable internal audit showed an agent operating at the level of a mid-level engineer end to end. Meta later described most code heading toward AI authorship rather than most engineering being replaced.", "tags": [ "coding", "labour", "agents" ], "sources": [ { "label": "Business Insider", "publisher": "Business Insider", "title": "Mark Zuckerberg Says AI Could Soon Do the Work of Midlevel Meta Coders", "url": "https://www.businessinsider.com/mark-zuckerberg-meta-ai-replace-engineers-coders-joe-rogan-podcast-2025-1", "date": "2025-01-11", "kind": "reporting", "establishes": "The quotation and its immediate context." }, { "label": "Engadget", "publisher": "Engadget", "title": "Zuckerberg predicts AI will write most of Meta’s code within 12 to 18 months", "url": "https://www.engadget.com/ai/mark-zuckerberg-predicts-ai-will-write-most-of-metas-code-within-12-to-18-months-213851646.html", "date": "2025-04-30", "kind": "reporting", "establishes": "The follow-on, narrower version of the same prediction." } ], "status": "verified" }, { "id": "altman-agents-join-workforce", "personId": "altman", "quote": "We believe that, in 2025, we may see the first AI agents “join the workforce” and materially change the output of companies.", "claim": "By the end of 2025, AI agents would be doing real work inside companies and would measurably change company output.", "domain": "work", "venue": "Blog post, “Reflections”", "madeOn": "2025-01-05", "horizon": { "kind": "exact", "end": "2025-12-31", "label": "By 31 December 2025" }, "falsifiable": true, "verdict": "contested", "verdictNote": "Agents were widely deployed in 2025 — coding agents, browser agents, enterprise “coworker” products — so the first clause held. The second, “materially change the output of companies”, is where the evidence splits: year-end reviews found little aggregate productivity signature, while vendors reported strong internal gains. Scored contested rather than validated.", "tags": [ "agents", "labour", "adoption" ], "featured": true, "sources": [ { "label": "Altman, Reflections", "publisher": "Sam Altman", "title": "Reflections", "url": "https://blog.samaltman.com/reflections", "date": "2025-01-05", "kind": "primary", "establishes": "The verbatim claim and its hedge (“may see”)." }, { "label": "The New Yorker", "publisher": "The New Yorker", "title": "Why A.I. Didn’t Transform Our Lives in 2025", "url": "https://www.newyorker.com/culture/2025-in-review/why-ai-didnt-transform-our-lives-in-2025", "date": "2025-12-27", "kind": "analysis", "establishes": "The sceptical year-in-review reading of the same prediction." }, { "label": "Cal Newport", "publisher": "Cal Newport", "title": "Why Didn’t AI “Join the Workforce” in 2025?", "url": "https://calnewport.com/why-didnt-ai-join-the-workforce-in-2025/", "date": "2026-01-06", "kind": "analysis", "establishes": "A one-year retrospective on the claim, with the original text quoted." } ], "status": "verified" }, { "id": "hinton-10-20-extinction", "personId": "hinton", "quote": "Not really, 10% to 20%.", "claim": "There is a 10–20% chance that AI leads to human extinction within roughly three decades.", "domain": "security", "venue": "BBC Radio 4 Today programme", "madeOn": "2024-12-27", "horizon": { "kind": "range", "start": "2024-12-27", "end": "2054-12-27", "label": "Next 30 years" }, "falsifiable": false, "statedProbability": 0.15, "verdict": "open", "verdictNote": "Hinton describes the figure as a gut feeling rather than a statistical analysis, and there is no agreed method for adjudicating extinction probability. Later revisions moved the timeline inward — “maybe 10 years, maybe less” — while keeping a ~10% figure.", "tags": [ "risk", "extinction", "probability" ], "sources": [ { "label": "The Guardian", "publisher": "The Guardian", "title": "“Godfather of AI” shortens odds of the technology wiping out humanity over next 30 years", "url": "https://www.theguardian.com/technology/2024/dec/27/godfather-of-ai-raises-odds-of-the-technology-wiping-out-humanity-over-next-30-years", "date": "2024-12-27", "kind": "reporting", "establishes": "The 10–20% figure and the exchange with Sajid Javid." }, { "label": "Business Insider", "publisher": "Business Insider", "title": "The “Godfather of AI” says it’s not “unreasonable” to believe there’s a 10% chance the technology will end humanity", "url": "https://www.businessinsider.com/geoffrey-hinton-godfather-of-ai-human-extinction-odds-2026-9", "date": "2026-09-10", "kind": "reporting", "establishes": "BBC Newsnight restatement: “maybe 10 years, maybe less.”" } ], "status": "verified" }, { "id": "bengio-window-not-ready", "personId": "bengio", "quote": "But if it’s five years, we’re not ready … because we don’t have methods to make sure that these systems will not harm people or will not turn against people … We don’t know how to do that.", "claim": "Within about five years, systems may be capable of harm we cannot prevent, because the safety methods do not exist yet.", "domain": "security", "venue": "CNBC interview", "madeOn": "2024-11-21", "horizon": { "kind": "exact", "end": "2029-11-21", "label": "By 2029" }, "falsifiable": true, "verdict": "open", "verdictNote": "Partly testable through the safety-tooling record rather than through an event: independent evaluators still report that no demonstrated method reliably controls systems substantially more capable than their overseers. The stronger claim — that capability outruns control — is a live research dispute.", "tags": [ "risk", "alignment", "governance" ], "sources": [ { "label": "CNBC", "publisher": "CNBC", "title": "AI systems could “turn against humans”: Yoshua Bengio warns of artificial intelligence risks", "url": "https://www.cnbc.com/2024/11/21/will-ai-replace-humans-yoshua-bengio-warns-of-artificial-intelligence-risks.html", "date": "2024-11-21", "kind": "reporting", "establishes": "The quotation and the five-year framing." }, { "label": "The Guardian / BBC Today", "publisher": "The Guardian", "title": "“Godfather of AI” shortens odds of the technology wiping out humanity over next 30 years", "url": "https://www.theguardian.com/technology/2024/dec/27/godfather-of-ai-raises-odds-of-the-technology-wiping-out-humanity-over-next-30-years", "date": "2024-12-27", "kind": "reporting", "establishes": "Hinton’s parallel 10–20% estimate, for comparison in the risk view." } ], "status": "verified" }, { "id": "su-500bn-2028", "personId": "su", "quote": "In the Data Center alone, we expect the AI accelerator TAM will grow at more than 60% annually to $500 billion in 2028.", "claim": "The data-centre AI accelerator market reaches $500bn by 2028.", "domain": "capital", "venue": "AMD Q3 2024 earnings call", "madeOn": "2024-10-29", "horizon": { "kind": "exact", "end": "2028-12-31", "label": "By 2028" }, "falsifiable": true, "verdict": "open", "verdictNote": "Too early to score, but the revision history is the story: AMD raised the same 2028 figure twice, then in 2026 projected $1.4tn by 2030 — and Su described the original $500bn as having become “conservative”. Market-size predictions from suppliers are simultaneously forecasts and sales targets.", "tags": [ "market size", "accelerators", "capex" ], "sources": [ { "label": "AMD earnings call (via Yahoo)", "publisher": "Yahoo Finance", "title": "AMD’s Lisa Su predicts AI GPU market will grow to $500 billion by 2028", "url": "https://finance.yahoo.com/news/amds-dr-lisa-su-predicts-172450644.html", "date": "2024-10-30", "kind": "reporting", "establishes": "The original 2028 claim and the 60% CAGR." }, { "label": "The Next Platform", "publisher": "The Next Platform", "title": "Picking Apart AMD’s AI Accelerator Forecasts", "url": "https://www.nextplatform.com/ai/2025/06/17/picking-apart-amds-ai-accelerator-forecasts-for-fun-and-budgets/1633312", "date": "2025-06-17", "kind": "analysis", "establishes": "Sceptical reading of the supplier-forecast incentive." } ], "status": "verified" }, { "id": "amodei-powerful-ai-2026", "personId": "amodei", "quote": "I think it could come as early as 2026, though there are also ways it could take much longer.", "claim": "“Powerful AI” — a system better than a Nobel laureate in most fields, able to run for days autonomously — could arrive as early as 2026.", "domain": "models", "venue": "Essay, “Machines of Loving Grace”", "madeOn": "2024-10-11", "horizon": { "kind": "range", "start": "2026-01-01", "end": "2027-12-31", "label": "2026 – 2027" }, "falsifiable": true, "verdict": "open", "verdictNote": "Anthropic’s own later policy submission narrowed this to “late 2026 or early 2027”. As of the snapshot no publicly evaluated system is accepted as beating Nobel laureates across most disciplines; the window is still open. This is the single most consequential dated claim in the catalogue.", "tags": [ "AGI", "timeline", "powerful AI" ], "featured": true, "sources": [ { "label": "Amodei, Machines of Loving Grace", "publisher": "Dario Amodei", "title": "Machines of Loving Grace", "url": "https://darioamodei.com/essay/machines-of-loving-grace", "date": "2024-10-11", "kind": "primary", "establishes": "The definition of “powerful AI” and the 2026 hedge." }, { "label": "Dwarkesh Podcast", "publisher": "Dwarkesh Patel", "title": "Dario Amodei — “We are near the end of the exponential”", "url": "https://www.dwarkesh.com/p/dario-amodei-2", "date": "2026-10-01", "kind": "primary", "establishes": "Amodei restates the hunch in 2026: “2026, maybe 2027… I wouldn’t be surprised if I’m off by a year or two.”" }, { "label": "Redwood Research", "publisher": "Redwood Research", "title": "What’s up with Anthropic predicting AGI by early 2027?", "url": "https://blog.redwoodresearch.org/p/whats-up-with-anthropic-predicting", "date": "2025-12-25", "kind": "analysis", "establishes": "Compares the essay’s hedge with Anthropic’s stronger OSTP submission." } ], "status": "verified" }, { "id": "altman-asi-thousand-days", "personId": "altman", "quote": "It is possible that we will have superintelligence in a few thousand days (!); it may take longer, but I’m confident we’ll get there.", "claim": "Superintelligence — AI vastly smarter than humans — may arrive within roughly 1,000 to 4,000 days of September 2024, i.e. between 2027 and 2035.", "domain": "models", "venue": "Essay, “The Intelligence Age”", "madeOn": "2024-09-23", "horizon": { "kind": "range", "start": "2027-06-20", "end": "2035-08-16", "label": "2027 – 2035" }, "falsifiable": true, "verdict": "open", "verdictNote": "The window runs to 2035, and Altman explicitly allowed that it “may take longer”, which makes the claim hard to falsify early. No system as of the snapshot is credibly described as superintelligent by the definition OpenAI has used (“vastly smarter than humans”). Tracked, unscored.", "tags": [ "AGI", "superintelligence", "timeline" ], "featured": true, "sources": [ { "label": "Altman, The Intelligence Age", "publisher": "Sam Altman", "title": "The Intelligence Age", "url": "https://ia.samaltman.com/", "date": "2024-09-23", "kind": "primary", "establishes": "The exact published sentence and its date." }, { "label": "Fortune", "publisher": "Fortune", "title": "Sam Altman says AI superintelligence could be just ‘a few thousand days’ away", "url": "https://fortune.com/2024/09/24/sam-altman-ai-superintelligence/", "date": "2024-09-24", "kind": "reporting", "establishes": "Contemporary coverage and the 2,000–5,000 day arithmetic." }, { "label": "Ars Technica", "publisher": "Ars Technica", "title": "OpenAI CEO: We may have AI superintelligence in “a few thousand days”", "url": "https://arstechnica.com/information-technology/2024/09/ai-superintelligence-looms-in-sam-altmans-new-essay-on-the-intelligence-age/", "date": "2024-09-24", "kind": "reporting", "establishes": "Independent reading of the vagueness of the unit." } ], "status": "verified" }, { "id": "narayanan-xrisk-numbers", "personId": "narayanan", "quote": "AI x-risk forecasts are far too unreliable to be useful for policy, and in fact highly misleading.", "claim": "Quantified AI existential-risk probabilities are not a sound basis for policy.", "domain": "security", "venue": "AI Snake Oil essay with Sayash Kapoor", "madeOn": "2024-07-26", "horizon": { "kind": "open", "start": "2024-07-26", "label": "Standing methodological claim" }, "falsifiable": true, "verdict": "contested", "verdictNote": "This is itself a claim about forecasting that can be checked by looking at forecast behaviour. The catalogue’s own evidence is mixed: stated probabilities differ by orders of magnitude between experts, which supports his point, while structured forecasting efforts have improved calibration on shorter-horizon questions.", "tags": [ "forecasting", "risk", "methodology" ], "sources": [ { "label": "Narayanan on X", "publisher": "X", "title": "On the deep dive into AI existential risk probability estimates", "url": "https://x.com/random_walker/status/1816816078577631331", "date": "2024-07-26", "kind": "primary", "establishes": "The authors’ own summary of the finding." }, { "label": "MIT Sloan", "publisher": "MIT Sloan", "title": "How to spot real value in AI — and avoid the snake oil", "url": "https://mitsloan.mit.edu/ideas-made-to-matter/how-to-spot-real-value-ai-and-avoid-snake-oil", "date": "2026-01-29", "kind": "analysis", "establishes": "The predictive-AI critique that grounds his forecasting scepticism." } ], "status": "verified" }, { "id": "son-asi-2035", "personId": "son", "quote": "Artificial superintelligence will be 10,000 times smarter than a human brain and will exist by 2035.", "claim": "By 2035, artificial superintelligence exists at roughly 10,000× human capability, with AGI arriving in the next three to five years.", "domain": "models", "venue": "SoftBank World / Future Investment Initiative remarks", "madeOn": "2024-06-21", "horizon": { "kind": "exact", "end": "2035-12-31", "label": "AGI 2027–2029 · ASI by 2035" }, "falsifiable": true, "verdict": "open", "verdictNote": "“10,000 times smarter” has no measurement scheme, so the claim as stated is not adjudicable. Its checkable core — AGI within three to five years of 2024, i.e. by 2029 — is still open. Son later said his public timeline was deliberately conservative and that he privately expected ASI in four years.", "tags": [ "ASI", "AGI", "capital" ], "featured": true, "sources": [ { "label": "CNBC", "publisher": "CNBC", "title": "SoftBank CEO says AI that is 10,000 times smarter than humans will come out in 10 years", "url": "https://www.cnbc.com/2024/06/21/softbank-ceo-predicts-ai-that-is-10000-times-smarter-than-humans-.html", "date": "2024-06-21", "kind": "reporting", "establishes": "The AGI-in-3–5-years and ASI-in-10-years structure." }, { "label": "Reuters", "publisher": "Reuters", "title": "SoftBank CEO talks up artificial super intelligence ambitions", "url": "https://www.reuters.com/technology/artificial-intelligence/softbank-ceo-talks-up-firms-ai-ambitions-2024-06-21/", "date": "2024-06-21", "kind": "reporting", "establishes": "The 10,000× framing at the shareholders’ meeting." }, { "label": "TIME", "publisher": "TIME", "title": "Masayoshi Son Is Betting It All on American AI", "url": "https://time.com/7380074/ai-masayoshi-son-japan-softbank/", "date": "2026-04-13", "kind": "reporting", "establishes": "The 10%-of-global-GDP framing for ASI plus robotics." } ], "status": "verified" }, { "id": "ng-agentic-workflows", "personId": "ng", "quote": "For the majority of businesses, focus on building applications using agentic workflows rather than solely scaling traditional AI. That’s where the greatest opportunity lies.", "claim": "The return on AI for most firms comes from agentic workflow engineering, not from bigger base models.", "domain": "work", "venue": "Snowflake Summit, Luminary Talk", "madeOn": "2024-06-03", "horizon": { "kind": "open", "start": "2024-06-03", "label": "Standing claim" }, "falsifiable": true, "verdict": "partial", "verdictNote": "Enterprise adoption data through 2026 supports the direction: value concentrated in workflow integration and evaluation harnesses, while pilots without process redesign failed at high rates. The stronger reading — that model scaling would not matter to firms — has not held.", "tags": [ "enterprise", "agents", "adoption" ], "sources": [ { "label": "Global Advisors", "publisher": "Global Advisors", "title": "Quote: Andrew Ng on agentic workflows", "url": "https://globaladvisors.biz/2025/07/06/quote-andrew-ng-ai-guru-3/", "date": "2025-07-06", "kind": "analysis", "establishes": "The quote and its Snowflake Summit provenance." }, { "label": "DeepLearning.AI", "publisher": "DeepLearning.AI", "title": "The AI Engineering Skills Map", "url": "https://www.deeplearning.ai/the-batch/the-ai-engineering-skills-map", "date": "2026-08-14", "kind": "primary", "establishes": "His 2026 operational view of where the work actually is." } ], "status": "verified" }, { "id": "musk-smarter-than-human-2025", "personId": "musk", "quote": "My guess is that we’ll have AI that is smarter than any one human probably around the end of next year.", "claim": "By the end of 2025, AI would be smarter than any single human being.", "domain": "models", "venue": "Interview with Nicolai Tangen, “In Good Company”", "madeOn": "2024-04-08", "horizon": { "kind": "exact", "end": "2025-12-31", "label": "End of 2025" }, "falsifiable": true, "verdict": "missed", "verdictNote": "2025 passed without a system generally accepted as smarter than any human across domains. The date then moved to end-2026 at Davos, while the definition stayed the same — a textbook moving-goalpost pattern. This record exists to keep the earlier date visible.", "tags": [ "AGI", "timeline", "moved goalposts" ], "featured": true, "sources": [ { "label": "Business Insider", "publisher": "Business Insider", "title": "Elon Musk Says AI Will Outsmart Humans by 2026", "url": "https://www.businessinsider.com/elon-musk-says-ai-will-outsmart-humans-by-2026-2024-4", "date": "2024-04-09", "kind": "reporting", "establishes": "The 2024 quote, made when “next year” meant 2025." }, { "label": "Fello AI tracker", "publisher": "Fello AI", "title": "When Will AGI Happen? Every Prediction, Tracked", "url": "https://felloai.com/when-will-agi-happen/", "date": "2026-10-06", "kind": "analysis", "establishes": "Documents the same quote with the 2025 date, and the 2026 revision." } ], "status": "verified" }, { "id": "leike-superalignment-2027", "personId": "leike", "quote": "Our goal is to solve the core technical challenges of superintelligence alignment in four years.", "claim": "The core technical challenges of aligning superintelligence would be solved by mid-2027, with 20% of OpenAI’s compute.", "domain": "security", "venue": "OpenAI, “Introducing Superalignment”; co-led with Ilya Sutskever", "madeOn": "2023-07-05", "horizon": { "kind": "exact", "end": "2027-07-05", "label": "By July 2027" }, "falsifiable": true, "verdict": "missed", "settledEarly": "Scored before the deadline by design: the commitment cannot be met because the team that made it was dissolved. Kept visible as an institutional accountability record rather than a live forecast.", "verdictNote": "The programme’s co-leads departed in 2024 and the team was dissolved; the commitment cannot now be delivered by the institution that made it, and no successor programme has claimed the deadline. This is scored as an institutional commitment that failed by attrition, not as a technical result — an important distinction for anyone reading lab forecasts as contracts.", "tags": [ "alignment", "commitment", "institutions" ], "featured": true, "sources": [ { "label": "OpenAI", "publisher": "OpenAI", "title": "Introducing Superalignment", "url": "https://openai.com/index/introducing-superalignment/", "date": "2023-07-05", "kind": "primary", "establishes": "The four-year goal and the 20% compute commitment." }, { "label": "AXRP", "publisher": "AXRP podcast", "title": "Superalignment with Jan Leike", "url": "https://axrp.net/episode/2023/07/27/episode-24-superalignment-with-jan-leike.html", "date": "2023-07-27", "kind": "primary", "establishes": "Leike’s own mid-2027 framing and 85% confidence." }, { "label": "Business Insider", "publisher": "Business Insider", "title": "It was their job to make sure humans are safe from OpenAI’s superintelligence. They just quit.", "url": "https://www.businessinsider.com/jan-leike-ilya-sutskever-resignations-superalignment-openai-superintelligence-safe-humanity-2024-5", "date": "2024-05-15", "kind": "reporting", "establishes": "The departures that ended the programme." } ], "status": "verified" }, { "id": "bender-meaning-gap", "personId": "bender", "quote": "An LM is a system for haphazardly stitching together sequences of linguistic forms it has observed in its vast training data, according to probabilistic information about how they combine, but without any reference to meaning: a stochastic parrot.", "claim": "Scaling language models will not produce grounded understanding, because form alone does not supply meaning.", "domain": "models", "venue": "“On the Dangers of Stochastic Parrots” (FAccT 2021)", "madeOn": "2021-03-01", "horizon": { "kind": "open", "start": "2021-03-01", "label": "Standing claim" }, "falsifiable": true, "verdict": "contested", "verdictNote": "The most durable disagreement in the catalogue. Evidence for her position: benchmark sensitivity shows performance resting on surface patterns. Evidence against: systems now pass expert exams and produce novel proofs, which many researchers treat as grounding. No agreed operational test for “understanding” exists — which is precisely her point.", "tags": [ "meaning", "LLM", "grounding" ], "featured": true, "sources": [ { "label": "Bender, Gebru, McMillan-Major, Shmitchell", "publisher": "ACM FAccT", "title": "On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?", "url": "https://dl.acm.org/doi/10.1145/3442188.3445922", "date": "2021-03-01", "kind": "primary", "establishes": "The paper and the definition quoted." }, { "label": "IEEE Spectrum", "publisher": "IEEE Spectrum", "title": "What Emily Bender Really Meant by “Stochastic Parrots”", "url": "https://spectrum.ieee.org/stochastic-parrot", "date": "2026-06-30", "kind": "primary", "establishes": "Her 2026 clarification of the claim’s scope." }, { "label": "Bender, Medium", "publisher": "Emily M. Bender", "title": "Stochastic Parrots: Frequently Unasked Questions", "url": "https://medium.com/@emilymenonbender/stochastic-parrots-frequently-unasked-questions-49c2e7d22d11", "date": "2026-05-12", "kind": "primary", "establishes": "Her own list of common misreadings." } ], "status": "verified" }, { "id": "brockman-position", "personId": "brockman", "claim": "The next decade of progress is decided by whoever can afford to run experiments at scale, not by whoever has the best idea.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "scaling", "labs" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "murati-position", "personId": "murati", "claim": "Useful intelligence will be built by teams that tune models to people, not by leaderboards.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "research", "products" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "pachocki-position", "personId": "pachocki", "claim": "Capability gains keep compounding from compute and data until the curve itself becomes the strategy.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "scaling", "research" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "mcgrew-position", "personId": "mcgrew", "claim": "Inference-time reasoning is the next axis of progress, and it will be paid for in silicon.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "reasoning", "compute" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "friar-position", "personId": "friar", "claim": "The capital required to stay at the frontier will keep rising faster than revenue, so financing is the real roadmap.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "financing" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "simo-position", "personId": "simo", "claim": "AI becomes real when it reaches ordinary workflows, and that means the consumer product is the strategy.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "products", "adoption" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "weil-position", "personId": "weil", "claim": "Every knowledge job will be restructured around a model that drafts first and a human that edits.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "work", "products" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "mchen-position", "personId": "mchen", "claim": "Models will start producing research results that their own creators cannot fully explain.", "domain": "science", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "science", "research" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "zaremba-position", "personId": "zaremba", "claim": "Robotics will follow language models: a general policy trained broadly beats hand-engineered control.", "domain": "robotics", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "robotics", "learning" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "madry-position", "personId": "madry", "claim": "Robustness is a research problem, not a compliance problem, and it will be solved or the deployments will not happen.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "safety", "robustness" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "olah-position", "personId": "olah", "claim": "Interpretability is the only way to audit a system nobody can read, and it will become a release requirement.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "interpretability", "safety" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "kaplan-position", "personId": "kaplan", "claim": "Scaling laws are the most reliable forecast in the field, and they still hold through the next order of magnitude.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "scaling", "forecasting" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "krieger-position", "personId": "krieger", "claim": "Agentic products will be judged by what they finish, not by what they answer.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "agents", "products" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "rao-position", "personId": "rao", "claim": "Enterprise demand, not consumer novelty, will decide which labs survive the decade.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "enterprise" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "clark-position", "personId": "clark", "claim": "Compute governance is the only lever that can be pulled quickly enough to matter.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "policy", "compute" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "askell-position", "personId": "askell", "claim": "Character training is a safety technique, and the industry will treat it as one.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "alignment", "character" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "hubinger-position", "personId": "hubinger", "claim": "Deceptive alignment is the failure mode that will define the decade.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "alignment", "risk" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "tom_brown-position", "personId": "tom_brown", "claim": "Pre-training scale still buys capability that no amount of post-training cleverness replaces.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "scaling", "pre-training" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "mccandlish-position", "personId": "mccandlish", "claim": "The cost of a given capability will fall by orders of magnitude, and that is what changes society.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "cost", "scaling" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "kavukcuoglu-position", "personId": "kavukcuoglu", "claim": "Multimodal models trained on everything will make single-purpose tools look like toys.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "multimodal", "research" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "legg-position", "personId": "legg", "claim": "AGI arrives through accumulation, and the accumulation is already visible in the benchmarks.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "agi", "forecasting" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "vinyals-position", "personId": "vinyals", "claim": "Scientific discovery will be the first domain where models outperform their trainers.", "domain": "science", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "science", "discovery" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "jeffdean-position", "personId": "jeffdean", "claim": "Specialised silicon plus sparse models is the only path to serving intelligence at internet scale.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "compute", "silicon" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "manyika-position", "personId": "manyika", "claim": "The productivity gains are real but the distribution is political, and policy will decide who gets them.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "policy", "labour" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "porat-position", "personId": "porat", "claim": "Infrastructure spending of this size is a bet on a decade, not a quarter.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "infrastructure" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "kurian-position", "personId": "kurian", "claim": "Enterprises will buy outcomes, not models, and the cloud that hosts the workflow wins.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "enterprise", "cloud" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "vahdat-position", "personId": "vahdat", "claim": "Power, not chips, becomes the binding constraint on training within a few years.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "compute", "energy" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "woodward-position", "personId": "woodward", "claim": "The winning AI product is invisible: it lives inside the tools people already use.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "products", "adoption" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "kilpatrick-position", "personId": "kilpatrick", "claim": "Cheap access to strong models is what turns an industry into an ecosystem.", "domain": "openness", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "developer", "access" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "shazeer-position", "personId": "shazeer", "claim": "Mixture-of-experts and scale are how you get frontier quality at serving cost.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "architecture", "scale" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "brin-position", "personId": "brin", "claim": "The models are already more capable than the products built on them, and that gap closes fast.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "products", "capability" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "jassy-position", "personId": "jassy", "claim": "Every application will be rebuilt around inference, and the bill will be paid by the application.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "cloud", "inference" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "garman-position", "personId": "garman", "claim": "Custom silicon is the difference between renting capacity and controlling cost.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "silicon", "cloud" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "prasad-position", "personId": "prasad", "claim": "Agents that act in the physical world will arrive through warehouses before homes.", "domain": "robotics", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "robotics", "agents" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "limp-position", "personId": "limp", "claim": "Voice and vision in the home will be normal before humanoid labour is.", "domain": "robotics", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "devices", "robotics" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "nadella-position", "personId": "nadella", "claim": "Every job will be redesigned around copilots, and the org chart is the real product surface.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "work", "enterprise" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "kevinscott-position", "personId": "kevinscott", "claim": "We will build datacentres at a scale that would have sounded absurd five years ago, and still fall short.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "compute", "infrastructure" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "rajeshjha-position", "personId": "rajeshjha", "claim": "The operating system becomes an agent that knows your work.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "agents", "products" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "chriscox-position", "personId": "chriscox", "claim": "Personal superintelligence will be the default assistant for billions, not a premium tier.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "products", "scale" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "bosworth-position", "personId": "bosworth", "claim": "Glasses and headsets will carry the assistant, because hands-free is the only interface that scales.", "domain": "robotics", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "devices", "interfaces" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "susuanli-position", "personId": "susuanli", "claim": "The market will eventually demand evidence that AI capex converts into revenue, and the answer arrives in stages.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "disclosure" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "babuschkin-position", "personId": "babuschkin", "claim": "Build the compute fast enough and capability follows on its own schedule.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "compute", "speed" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "gregyang-position", "personId": "gregyang", "claim": "Mathematical theory of scale will let us predict capability before we spend the money.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "theory", "scaling" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "intrator-position", "personId": "intrator", "claim": "Specialised GPU clouds will outgrow the general clouds for training workloads.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "compute", "cloud" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "schroepfer-position", "personId": "schroepfer", "claim": "Climate and AI capital converge: the winners will be judged on energy, not just models.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "climate" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "natfriedman-position", "personId": "natfriedman", "claim": "Open-weight models will become the default substrate for most software.", "domain": "openness", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "open weights", "investment" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "gross-position", "personId": "gross", "claim": "The next great companies are being founded by people who never write code by hand.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "building" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "hoffman-position", "personId": "hoffman", "claim": "AI is a general-purpose amplifier of human agency, and the returns go to whoever adopts it first.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "work", "adoption" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "khosla-position", "personId": "khosla", "claim": "AI will do the work of most doctors and lawyers within a generation, and that is a public good.", "domain": "science", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "science", "labour" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "andreessen-position", "personId": "andreessen", "claim": "Software eats the world, and now software builds the software.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "software" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "horowitz-position", "personId": "horowitz", "claim": "The AI build-out is a capital cycle, and capital cycles end.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "markets" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "thiel-position", "personId": "thiel", "claim": "The bottleneck is not intelligence, it is the will to deploy it into regulated industries.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "regulation" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "palihapitiya-position", "personId": "palihapitiya", "claim": "Public markets will reprice AI infrastructure sharply before the decade is out.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "markets", "capital" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "gerstner-position", "personId": "gerstner", "claim": "We are early in the capex cycle and late in the hype cycle.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "markets", "capex" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "gavinbaker-position", "personId": "gavinbaker", "claim": "Memory bandwidth, not FLOPs, will determine who can serve the next generation of models.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "compute", "silicon" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "sarahguo-position", "personId": "sarahguo", "claim": "Every repeatable service business is a candidate to become software with a model inside it.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "services", "software" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "eladgil-position", "personId": "eladgil", "claim": "The consolidation of AI startups into a handful of platforms will be faster than the cloud transition.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "consolidation", "capital" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "srinivas-position", "personId": "srinivas", "claim": "Answers, not links, become the interface to the web.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "search", "products" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "harrisonchase-position", "personId": "harrisonchase", "claim": "Agent infrastructure is the plumbing of the next decade of software.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "agents", "tools" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "amjadmasad-position", "personId": "amjadmasad", "claim": "Natural language becomes the programming language, and the number of software creators multiplies.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "coding", "access" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "guillermorauch-position", "personId": "guillermorauch", "claim": "The web becomes generated per user, per request.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "web", "products" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "alexandrwang-position", "personId": "alexandrwang", "claim": "Data, not algorithms, is the moat that compounds.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "data", "enterprise" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "lucyguo-position", "personId": "lucyguo", "claim": "Creator and human attention markets get repriced by generated media.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "media", "markets" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "dylanfield-position", "personId": "dylanfield", "claim": "Design tools become collaborators, and craft becomes taste plus direction.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "design", "tools" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "brettaylor-position", "personId": "brettaylor", "claim": "Customer service is the first industry fully reshaped by agents, because the feedback loop is immediate.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "agents", "service" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "claybavor-position", "personId": "claybavor", "claim": "An AI agent with a budget and a goal replaces a queue of tickets.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "agents", "enterprise" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "alidhodsi-position", "personId": "alidhodsi", "claim": "Enterprise data plus models is where the durable margin lives.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "enterprise", "data" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "mateizaharia-position", "personId": "mateizaharia", "claim": "Open research infrastructure will keep pace with private labs on applied work.", "domain": "science", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "research", "open" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "ionstoica-position", "personId": "ionstoica", "claim": "Serving efficiency improvements will decide whether AI is affordable at scale.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "serving", "efficiency" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "naveenrao-position", "personId": "naveenrao", "claim": "Accelerator diversity is necessary, because one vendor cannot supply the world.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "silicon", "supply" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "mensch-position", "personId": "mensch", "claim": "Open and sovereign models will be the European answer to concentrated capability.", "domain": "openness", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "open weights", "sovereignty" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "lample-position", "personId": "lample", "claim": "Efficient training on curated data closes most of the gap with brute force.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "efficiency", "research" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "dettmers-position", "personId": "dettmers", "claim": "Quantisation and memory efficiency matter more than a few points of benchmark score.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "efficiency", "hardware" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "thomaswolf-position", "personId": "thomaswolf", "claim": "The open ecosystem will ship most of the world's deployed models.", "domain": "openness", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "open weights", "ecosystem" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "chaumond-position", "personId": "chaumond", "claim": "Distribution of models, not training, becomes the platform layer.", "domain": "openness", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "distribution", "platform" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "soumith-position", "personId": "soumith", "claim": "Frameworks decide which hardware wins, which is why open tooling is strategic.", "domain": "openness", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "tooling", "hardware" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "percyliang-position", "personId": "percyliang", "claim": "Evaluation is infrastructure: without it, capability claims are marketing.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "evaluation", "research" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "christopherre-position", "personId": "christopherre", "claim": "Weak supervision and data programming will prove more valuable than another architecture paper.", "domain": "science", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "data", "research" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "chelseafinn-position", "personId": "chelseafinn", "claim": "Robot learning generalises when the data collection loop is designed, not bought.", "domain": "robotics", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "robotics", "learning" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "hendrycks-position", "personId": "hendrycks", "claim": "Safety must be treated as a global public good, with the same urgency as nuclear security.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "safety", "risk" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "stuartrussell-position", "personId": "stuartrussell", "claim": "The standard model of intelligence is wrong, and building it at scale is the risk.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "alignment", "theory" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "dawnsong-position", "personId": "dawnsong", "claim": "Security of AI systems becomes the limiting factor on deployment in finance and health.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "security", "systems" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "ancadragan-position", "personId": "ancadragan", "claim": "Value alignment is an engineering discipline with testable specifications.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "alignment", "testing" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "steinhardt-position", "personId": "steinhardt", "claim": "Forecasting and measurement will expose how much of the safety story is narrative.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "evaluation", "forecasting" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "aaronson-position", "personId": "aaronson", "claim": "Verification and watermarking of machine output will become a public utility.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "verification", "watermarks" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "tegmark-position", "personId": "tegmark", "claim": "The only stable outcome is a treaty regime for frontier capability.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "policy", "risk" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "koller-position", "personId": "koller", "claim": "AI-designed experiments compress drug discovery from decades to years.", "domain": "science", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "science", "biotech" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "barzilay-position", "personId": "barzilay", "claim": "Clinical models will be validated on outcomes rather than benchmarks, and that changes medicine.", "domain": "science", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "health", "validation" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "topol-position", "personId": "topol", "claim": "Medicine will be practised with a model in the loop, and the clinician's judgement becomes the product.", "domain": "science", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "health", "clinical" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "garymarcus-position", "personId": "garymarcus", "claim": "Scaling alone will not deliver reasoning, and the field will be forced to admit it.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "reasoning", "critique" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "melanie-position", "personId": "melanie", "claim": "Abstraction and analogy are the missing pieces, and they are not a data problem.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "cognition", "critique" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "katecrawford-position", "personId": "katecrawford", "claim": "Every model is an extractive industry with a supply chain of mines and labour.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "political economy", "compute" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "joyb-position", "personId": "joyb", "claim": "Biased systems at scale produce a new class of harm that audits must be empowered to stop.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "bias", "accountability" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "luccioni-position", "personId": "luccioni", "claim": "The energy and carbon of inference will be published like fuel economy, because it will be regulated.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "energy", "transparency" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "whittaker-position", "personId": "whittaker", "claim": "Surveillance business models are the default of AI deployment, and privacy must be structural.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "privacy", "power" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "rumman-position", "personId": "rumman", "claim": "Red-teaming becomes a licensed discipline with liability attached.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "audits", "safety" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "rachelthomas-position", "personId": "rachelthomas", "claim": "The deskilling of knowledge work will be the defining labour story of the decade.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "labour", "education" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "jeremyhoward-position", "personId": "jeremyhoward", "claim": "Top-down AI adoption fails; practical training inside the job is what works.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "education", "adoption" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "brynjolfsson-position", "personId": "brynjolfsson", "claim": "Measured productivity will rise sharply, later than the hype and faster than the sceptics expect.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "productivity", "economics" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "acemoglu-position", "personId": "acemoglu", "claim": "Most near-term AI gains will be modest, and the distributional effects are the real issue.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "economics", "labour" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "simonjohnson-position", "personId": "simonjohnson", "claim": "Concentration of AI capital threatens the political economy of democracies.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "power" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "tylercowen-position", "personId": "tylercowen", "claim": "AI accelerates growth unevenly, and the bottleneck becomes human coordination.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "growth", "economics" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "noahsmith-position", "personId": "noahsmith", "claim": "The job apocalypse is late, and the wage effects are showing up first for junior work.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "labour", "economics" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "korinek-position", "personId": "korinek", "claim": "Automation of cognitive work requires new institutions for income, not just retraining.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "policy", "labour" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "furman-position", "personId": "furman", "claim": "The fiscal effect of AI is the largest unmodelled variable in any budget forecast.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "policy", "fiscal" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "summers-position", "personId": "summers", "claim": "The investment boom will end in a correction before it ends in a plateau.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "markets", "capital" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "raimondo-position", "personId": "raimondo", "claim": "Export controls are the sharpest tool in the AI competition, and they will be widened.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "export controls", "policy" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "linakhan-position", "personId": "linakhan", "claim": "Conduct remedies will not fix AI concentration; structural separation will be required.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "competition", "policy" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "kanter-position", "personId": "kanter", "claim": "Vertical integration in AI infrastructure is the antitrust case of the decade.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "competition", "law" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "vestager-position", "personId": "vestager", "claim": "Europe will regulate first and compete later, and it will regret the order.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "regulation", "eu" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "breton-position", "personId": "breton", "claim": "The AI Act is a product-safety law, and enforcement will be export-facing.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "regulation", "eu" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "von_der_leyen-position", "personId": "von_der_leyen", "claim": "Compute investment is a sovereignty question, not an industrial policy question.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "sovereignty", "eu" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "kratsios-position", "personId": "kratsios", "claim": "The state will fund compute the way it once funded highways.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "government", "compute" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "sacks-position", "personId": "sacks", "claim": "Deregulation of deployment is the fastest route to capability advantage.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "deregulation", "policy" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "schumer-position", "personId": "schumer", "claim": "AI legislation will arrive as a package, after an election, and it will be narrower than the hearings suggest.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "legislation", "us" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "nickclegg-position", "personId": "nickclegg", "claim": "Regulation will converge on transparency duties because capability rules are unenforceable.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "transparency", "policy" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "bradsmith-position", "personId": "bradsmith", "claim": "Licensing and liability frameworks, not moratoria, will govern frontier models.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "law", "policy" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "kentwalker-position", "personId": "kentwalker", "claim": "Security cooperation between states and labs becomes routine and mostly invisible.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "security", "policy" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "makanju-position", "personId": "makanju", "claim": "Frontier labs will become diplomatic actors whether or not anyone licenses them to be.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "diplomacy", "policy" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "lehane-position", "personId": "lehane", "claim": "The political fight over AI is really a fight over who owns the story of progress.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "politics", "narrative" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "aschenbrenner-position", "personId": "aschenbrenner", "claim": "Automated AI research is the steepest capability curve in history, and the window to act is a few years.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "agi", "risk" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "yudkowsky-position", "personId": "yudkowsky", "claim": "Anything short of a moratorium on frontier capability is a decision to accept extinction risk.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "risk", "moratorium" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "soares-position", "personId": "soares", "claim": "Alignment is not a product feature, and treating it as one is the central mistake.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "alignment", "risk" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "zvi-position", "personId": "zvi", "claim": "The gap between stated safety practice and actual practice is the story to track.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "analysis", "critique" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "kokotajlo-position", "personId": "kokotajlo", "claim": "Superintelligent systems arrive by the end of the decade, well before institutions adapt.", "domain": "models", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "agi", "forecasting" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "deanball-position", "personId": "deanball", "claim": "Regulatory capture by incumbents is the likeliest safety failure of the next five years.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "policy", "competition" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "helentoner-position", "personId": "helentoner", "claim": "Evaluation capacity in government is the bottleneck for any serious oversight.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "oversight", "evaluation" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "tinocuellar-position", "personId": "tinocuellar", "claim": "Evidence institutions must be funded at the scale of the technology they study.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "evidence", "institutions" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "kaifulee-position", "personId": "kaifulee", "claim": "AI will displace routine cognitive work faster in services-heavy economies than in manufacturing ones.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "work", "china" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "wenfeng-position", "personId": "wenfeng", "claim": "Constraint forces efficiency, and efficiency will make open weights competitive.", "domain": "openness", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "open weights", "efficiency" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "zhangyiming-position", "personId": "zhangyiming", "claim": "Recommendation systems and language models converge into one demand engine.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "recommendation", "scale" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "liangrubo-position", "personId": "liangrubo", "claim": "China's advantage is deployment density, not frontier spend.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "deployment", "capital" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "wangxing-position", "personId": "wangxing", "claim": "Robotics and AI converge in last-mile logistics long before they converge in the home.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "robotics", "logistics" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "renzhengfei-position", "personId": "renzhengfei", "claim": "Domestic silicon supply chains will close the gap faster than export controls expect.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "silicon", "sovereignty" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "mengwanzhou-position", "personId": "mengwanzhou", "claim": "Compute sovereignty is national infrastructure, planned in decades.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "sovereignty", "infrastructure" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "leijun-position", "personId": "leijun", "claim": "AI moves into devices and vehicles before it moves into offices.", "domain": "robotics", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "devices", "robotics" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "luofuli-position", "personId": "luofuli", "claim": "Open technical reports will keep Chinese labs in the global research conversation.", "domain": "openness", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "research", "open" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "ponyma-position", "personId": "ponyma", "claim": "Applications, not models, are where the margin sits, and distribution decides.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "products", "distribution" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "liuqingfeng-position", "personId": "liuqingfeng", "claim": "Speech and language interfaces will reach hundreds of millions through education and government.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "access", "government" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "ericschmidt-position", "personId": "ericschmidt", "claim": "The United States must treat compute like a strategic reserve and fund it accordingly.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "strategy", "compute" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "renehaas-position", "personId": "renehaas", "claim": "Power efficiency, not peak performance, decides what gets deployed at the edge.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "energy", "chips" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "cristianoamon-position", "personId": "cristianoamon", "claim": "On-device inference is the answer to both cost and privacy.", "domain": "robotics", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "edge", "inference" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "lipbutan-position", "personId": "lipbutan", "claim": "Advanced packaging and foundry capacity decide who can build the accelerators.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "foundry", "chips" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "sanjaymehrotra-position", "personId": "sanjaymehrotra", "claim": "Memory is the constraint nobody models, and it will gate the next generation.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "memory", "supply" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "hocktan-position", "personId": "hocktan", "claim": "Custom accelerators will carry a third of training workloads.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "custom silicon", "networking" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "michaeldell-position", "personId": "michaeldell", "claim": "The enterprise AI stack is a hardware business with a software margin.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "hardware", "enterprise" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "antonioneri-position", "personId": "antonioneri", "claim": "Liquid cooling and power delivery are the unglamorous bottlenecks on every plan.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "datacentres", "cooling" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "chuckrobbins-position", "personId": "chuckrobbins", "claim": "Networking capacity is the silent limit on cluster scale.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "networking", "scale" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "arvindkrishna-position", "personId": "arvindkrishna", "claim": "Governed enterprise AI beats ungoverned frontier access for regulated industries.", "domain": "work", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "governance", "enterprise" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "donaldtrump-position", "personId": "donaldtrump", "claim": "The state will treat AI compute as a national asset and negotiate access the way it negotiates energy.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "compute", "strategy" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "jdvance-position", "personId": "jdvance", "claim": "Open-weight models are a check on concentrated power, and the state has an interest in protecting them.", "domain": "openness", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "open weights", "competition" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "marcorubio-position", "personId": "marcorubio", "claim": "Export controls and visa policy are the real levers of AI competition, not model releases.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "export controls", "diplomacy" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "scottbessent-position", "personId": "scottbessent", "claim": "Fiscal capacity, not model capability, sets the ceiling on a national AI programme.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "fiscal" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "howardlutnick-position", "personId": "howardlutnick", "claim": "Permitting and procurement will move faster than any frontier model release.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "industry", "procurement" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "chriswright-position", "personId": "chriswright", "claim": "Baseload power and permitting reform decide which country can actually train at scale.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "energy", "permitting" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "xi-position", "personId": "xi", "claim": "Compute sovereignty is a precondition of technological self-reliance.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "sovereignty", "china" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "liqiang-position", "personId": "liqiang", "claim": "Applied AI in industry and public services is the national priority, ahead of frontier research.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "deployment", "china" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "modi-position", "personId": "modi", "claim": "AI adoption at population scale is an infrastructure project, not a laboratory project.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "adoption", "india" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "vaishnaw-position", "personId": "vaishnaw", "claim": "Open models and public compute are how a country of a billion people adopts AI.", "domain": "openness", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "public compute", "open weights" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "starmer-position", "personId": "starmer", "claim": "AI inside public services is the fastest visible return on a national AI strategy.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "public services", "uk" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "reeves-position", "personId": "reeves", "claim": "The fiscal gain from AI arrives later than the political cycle that has to fund it.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "fiscal", "uk" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "kyle-position", "personId": "kyle", "claim": "Sovereign compute and an independent safety institute are the two halves of national capacity.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "compute", "safety" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "macron-position", "personId": "macron", "claim": "Europe competes by building its own capacity, not by regulating someone else's.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "sovereignty", "eu" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "merz-position", "personId": "merz", "claim": "Industrial AI adoption decides whether an economy keeps its manufacturing base.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "industry", "germany" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "meloni-position", "personId": "meloni", "claim": "Control of data and models is a security question before it is an economic one.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "security", "italy" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "sanchez-position", "personId": "sanchez", "claim": "AI governance will be written by whoever deploys it in public services first.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "public services", "spain" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "sejourne-position", "personId": "sejourne", "claim": "Simplifying the rulebook is the industrial policy Europe has been missing.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "regulation", "eu" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "lagarde-position", "personId": "lagarde", "claim": "Productivity from AI will show up in wages before it shows up in prices.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "monetary policy", "productivity" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "carney-position", "personId": "carney", "claim": "AI capital will flow to jurisdictions with cheap power and predictable permitting.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "canada" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "albanese-position", "personId": "albanese", "claim": "Critical minerals and compute are the same strategic question.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "minerals", "australia" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "lula-position", "personId": "lula", "claim": "AI governance must not become a new form of dependency for the global south.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "governance", "brazil" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "ramaphosa-position", "personId": "ramaphosa", "claim": "Data and compute sovereignty are prerequisites for an African AI economy.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "sovereignty", "africa" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "mbs-position", "personId": "mbs", "claim": "Cheap energy plus sovereign capital is a durable advantage in the AI build-out.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "energy", "state capital" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "mbz-position", "personId": "mbz", "claim": "State capital will anchor AI capacity wherever private capital is too slow.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "state capital", "uae" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "prabowo-position", "personId": "prabowo", "claim": "AI adoption in education and health is how a large population participates in the gains.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "adoption", "indonesia" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "wong-position", "personId": "wong", "claim": "Small states survive by being early, trusted and technically deep.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "strategy", "singapore" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "anwar-position", "personId": "anwar", "claim": "Neutrality plus compute investment is a viable position between the blocs.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "strategy", "malaysia" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "ishiba-position", "personId": "ishiba", "claim": "Semiconductor and AI supply chains need allied coordination, not autarky.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "supply chain", "japan" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "leejm-position", "personId": "leejm", "claim": "Memory and foundry capacity are national security assets.", "domain": "compute", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "memory", "korea" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "rutte-position", "personId": "rutte", "claim": "Allied coordination on AI security will matter more than any single national act.", "domain": "security", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "security", "alliances" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "guterres-position", "personId": "guterres", "claim": "Without a shared floor of rules, AI becomes a driver of inequality between states.", "domain": "policy", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "governance", "inequality" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "georgieva-position", "personId": "georgieva", "claim": "AI will widen the gap between economies that can adopt it and those that cannot.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "capital", "development" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "banga-position", "personId": "banga", "claim": "Access to compute, not access to papers, is the development bottleneck.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "development", "compute" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" }, { "id": "okonjoiweala-position", "personId": "okonjoiweala", "claim": "Trade rules for digital services will be rewritten around AI within a decade.", "domain": "capital", "venue": "Editorial summary", "horizon": { "kind": "open", "label": "Standing position" }, "falsifiable": true, "verdict": "unrated", "verdictNote": "Editorial summary of a publicly argued position, recorded without a verbatim quote, date or citation. It is not graded and is excluded from every verdict count.", "tags": [ "trade", "rules" ], "sources": [], "status": "editorial", "asOf": "2026-10-26" } ], "license": "CC BY 4.0 for original text; linked sources remain the property of their publishers.", "editorial": { "batch": "editorial-2026-10-26 + editorial-governments-2026-10-26", "asOf": "2026-10-26", "note": "Editorial roster additions. Each entry is OUR SUMMARY of a publicly argued position, recorded without a verbatim quote, a date or a citation, so none of them is graded. They exist to widen the record, and are labelled 'editorial' everywhere in the UI. Promote an entry only by replacing it with a sourced, graded record. Editorial additions for heads of government, ministers and multilateral leaders. Each entry is OUR SUMMARY of a publicly argued position, recorded without a verbatim quote, a date or a citation, so none of them is graded. Offices are as of the editorial date and change often: re-verify the office before promoting any record to cited. Nothing here is a quotation and nothing is attributed as one.", "entries": 184, "batches": [ { "batch": "editorial-2026-10-26", "asOf": "2026-10-26", "entries": 149 }, { "batch": "editorial-governments-2026-10-26", "asOf": "2026-10-26", "entries": 35 } ] } }