Worked examples · Article 46 of 60
Worked example: Starmer's ten million
A government target with a mechanism attached, which makes it unusually checkable.
A target for the number of people trained is one of the better kinds of state claim, because it has an owner, a programme and a count. The horizon is explicit, and the distance travelled is published.
Whether the training produces anything is a separate question, and the ledger keeps it separate.
Keir Starmer
Prime Minister · UK Government
AI inside public services is the fastest visible return on a national AI strategy.
1 claims recorded · 0 graded · 1 still open
Photograph: Simon Dawson / No 10 Downing Street — Open Government Licence v3.0
“We have begun the biggest skills drive in a generation – having worked with industry to deliver over one million AI courses towards our goal of upskilling 10 million workers by 2030.”
| Claim | Horizon | Verdict |
|---|---|---|
| <div><em>“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.”</em><br /><span class="tiny">Arvind Narayanan — Fortune interview on journalism and AI, 2026-10-06</span></div> | Open-ended, near term | <span class="verdict" data-v="open">Open</span> |
| <div><em>“Doomsday discussions distract from harms already occurring.”</em><br /><span class="tiny">Emily M. Bender — Bloomberg Technology interview, 2026-10-06</span></div> | Standing claim | <span class="verdict" data-v="open">Open</span> |
| <div><em>“Instead of discussing faulty, unreliable systems, tech leaders are marketing their systems as all-knowing, superpowerful, superintelligent, and lobbying for self-policing and…”</em><br /><span class="tiny">Timnit Gebru — Democracy Now! interview after the Right Livelihood Award, 2026-10-01</span></div> | Current policy cycle | <span class="verdict" data-v="contested">Contested</span> |
Figures computed from docs/predictions/ledger.json at build time. Every claim quoted on this page carries its own sources in the ledger.