Scores by kind of claim
Each kind of claim is scored on its own, out of 100: the share that held up, with the range it likely falls in. Fewer resolved claims give a wider range.
-
Factual claims
Statements about the present or the past, checked against the record: did they hold up?
80 72 88 -
Predictions
Statements about events outside the person's control, checked once they fell due: did they come true?
70 62 78 -
Promises
Commitments to the person's or organisation's own future action, with something to deliver: was it delivered?
Insufficient record: 2 promises resolved, 5 needed for a score.
All types combined: 75 (likely 69 to 80, high confidence), from 79 claims: 35 factual claims, 44 predictions. It mixes kinds of claim in different proportions for different people, so compare people kind by kind.
Scott Alexander's claims
35 claims
“Physicians For AI Safety wishes to advertise that they exist, are interested in talking to any physicians interested in AI safety, and have an open letter that such people can sign.”
Scott Alexander, ACX author 5 October 2026 Open Thread 454 (Astral Codex Ten)
Physicians For AI Safety exists, wants to talk to physicians interested in AI safety, and has an open letter such people can sign.
Mostly true True in substance, wrong in a detail.
The second judge would not resolve this claim; the verdict shown is the first judge's.
What would count as true or false
The claim has three parts, each judged as of 2026-10-05: (1) an organisation called Physicians For AI Safety existed; (2) it invited contact from physicians interested in AI safety; (3) it had an open letter that physicians could sign. TRUE: the organisation's website or another dated record shows that it existed, offered a way for physicians to get in touch, and hosted an open letter accepting signatures at or around that date. MOSTLY TRUE: the organisation and the open letter both existed, but a minor detail was off (for example signing was by request rather than an open form, or the name was slightly different). MIXED: the organisation existed, but there was no signable open letter, or the letter was closed to new signatures. MOSTLY FALSE: only a precursor or informal group existed, with no letter. FALSE: dated records show no such organisation existed at that time.
Evidence
- Physicians' Statement on AI Risk Physicians for AI Safety (organisation's own website) · 1 November 2024
“Physicians for AI Safety is an independent initiative started in September 2026 by three physicians from the United States, Germany and the United Kingdom.”
- Physicians' Statement on AI Risk Physicians for AI Safety (organisation's own website) · 1 November 2024
“You will receive an email to confirm your signature. Once confirmed and reviewed, your name is added to the list.”
- Physicians For AI Safety blogforiowa.com · 4 October 2026
“If you are willing, signing takes a minute at physiciansforaisafety.org.”
- Physicians for AI Safety: we must prevent harm from advanced AI BJGP Life · 29 October 2026 Own record
“We ask readers to sign it at www.physiciansforaisafety.org and to share it with colleagues.”
Reasoning
The organisation's website says Physicians for AI Safety was started in September 2026 by three physicians (E1), and it describes a signing process for its statement: signers confirm by email and their names are added after review (E2). A blog post dated 2026-10-04, independent of the subject, says signing takes a minute at physiciansforaisafety.org (E3), and a later article asks readers to sign there (E4). Together these indicate the organisation and a signable open letter existed around 2026-10-05. However, none of the quotes shows a specific way for physicians to get in touch beyond signing, and signatures are reviewed before they appear, so the claim is mostly true rather than fully true.
Cite this claim
Scott Alexander, “Physicians For AI Safety exists, wants to talk to physicians interested in AI safety, and has an open letter such people can sign.” (said 5 October 2026). Verdict: mostly true. Reliability API, claim c_c804ed2adff6, method 0.1.0+8bea457b. https://epistacklabs.com/claims/c_c804ed2adff6
Claims without a verdict
Is a verdict here wrong, or do you speak for Scott Alexander? Reply on the claim's own page, with the form at its foot (how replies work). Verdicts are made automatically; how verdicts and scores are made.