Worked examples · Article 53 of 60
Worked example: a claim that was missed
What a failure looks like when the record is written before the outcome is known.
The missed column is small, and that is a feature of an archive this young rather than a sign of accuracy. A miss is recorded the same way as a hit: the claim, the horizon, the evidence and a note that says what happened.
Reading the misses is the best way to understand the grading standard, because a standard that has never produced a failure has never been tested.
| Claim | Horizon | Verdict |
|---|---|---|
| <div><em>“My guess is that we’ll have AI that is smarter than any one human probably around the end of next year.”</em><br /><span class="tiny">Elon Musk — Interview with Nicolai Tangen, “In Good Company”, 2024-04-08</span></div> | End of 2025 | <span class="verdict" data-v="missed">Missed</span> |
| <div><em>“Our goal is to solve the core technical challenges of superintelligence alignment in four years.”</em><br /><span class="tiny">Jan Leike — OpenAI, “Introducing Superalignment”; co-led with Ilya Sutskever, 2023-07-05</span></div> | By July 2027 | <span class="verdict" data-v="missed">Missed</span> |
| Claim | Horizon | Verdict |
|---|---|---|
| <div><em>“Breathless claims about AGI and new capabilities fall apart pretty quickly under scrutiny.”</em><br /><span class="tiny">Timnit Gebru — MIT Technology Review, with Emily M. Bender, 2026-09-22</span></div> | 2026 claim cycle | <span class="verdict" data-v="partial">Partly borne out</span> |
| <div><em>“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.”</em><br /><span class="tiny">Aidan Gomez — CNBC “The Tech Download”, 2026-09-14</span></div> | Present tense | <span class="verdict" data-v="partial">Partly borne out</span> |
| <div><em>“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…”</em><br /><span class="tiny">Beth Barnes — METR paper, “Measuring AI Ability to Complete Long Tasks”, 2025-03-19</span></div> | Continuous trend | <span class="verdict" data-v="partial">Partly borne out</span> |
| <div><em>“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.”</em><br /><span class="tiny">Andrew Ng — Snowflake Summit, Luminary Talk, 2024-06-03</span></div> | Standing claim | <span class="verdict" data-v="partial">Partly borne out</span> |
Figures computed from docs/predictions/ledger.json at build time. Every claim quoted on this page carries its own sources in the ledger.