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?
83 78 88 -
Predictions
Statements about events outside the person's control, checked once they fell due: did they come true?
Insufficient record: 2 predictions resolved, 5 needed for a score.
-
Promises
Commitments to the person's or organisation's own future action, with something to deliver: was it delivered?
Insufficient record: 0 promises resolved, 5 needed for a score.
All types combined: 83 (likely 78 to 88, high confidence), from 86 claims: 86 factual claims. It mixes kinds of claim in different proportions for different people, so compare people kind by kind.
Daniela Amodei's claims
38 claims
“I think part of what we're seeing and the potential shorter term impacts of some of the AI safety research that we do is that different sized models exhibit different safety issues.”
Daniela Amodei, Anthropic president 2022 Podcast Transcript: Daniela and Dario Amodei on Anthropic (LessWrong)
Daniela Amodei said she thinks part of what Anthropic's AI safety research is seeing is that different sized models exhibit different safety issues.
Mostly true True in substance, wrong in a detail.
The second judge said resolved true; the verdict shown is the first judge's.
Said for an organisation they lead or work for; it counts on both records.
What would count as true or false
The claim is that Daniela Amodei said this. True if a recording or transcript of the 2022 interview shows her saying, in substance, that part of what Anthropic's safety research is finding is that models of different sizes show different safety issues. Mostly true if the wording differs but the meaning is the same. Mixed if she said something related but notably different, such as only about capabilities. Mostly false if the statement was made by someone else or its meaning was substantially changed. False if no such statement appears in the source, or if the source shows she said the opposite.
Evidence
- Daniela and Dario Amodei on Anthropic - Future of Life Institute Future of Life Institute · 16 November 2022
“Daniela Amodei is Co-Founder and President of Anthropic. She was previously at Stripe and OpenAI, and has also served as a congressional staffer.”
- Daniela and Dario Amodei on Anthropic · Zencastr Video Zencastr (hosts a transcript of the Future of Life Institute podcast) · 1 May 2024
“I think part of what we're seeing and the sort of potential shorter term impacts of some of the AI safety research that we do is that different sized models exhibit different safety issues, right?”
- Daniela and Dario Amodei on Anthropic · Zencastr Video Zencastr (hosts a transcript of the Future of Life Institute podcast) · 1 May 2024
“Yeah, I think one point Daniella made that's really important is this sudden emergence or change.”
Reasoning
A transcript of the Future of Life Institute podcast with Daniela and Dario Amodei contains the quoted line almost word for word: "different sized models exhibit different safety issues" [E4]. The quoted transcript excerpt in E4 does not itself name the speaker. However, the next speaker says "one point Daniella made that's really important" [E5], which supports attributing the line to Daniela Amodei. The FLI episode page dates the interview to 2022 and identifies her as Anthropic's co-founder and president [E3]. The wording differs slightly from the quote ("the sort of potential shorter term impacts") but the meaning is the same, so under the criteria this resolves as mostly true; attribution to Daniela Amodei rests on the episode featuring her [E3] and another speaker in the same transcript referring to a point 'Daniella' made [E5].
Cite this claim
Daniela Amodei, “Daniela Amodei said she thinks part of what Anthropic's AI safety research is seeing is that different sized models exhibit different safety issues.” (said 2022). Verdict: mostly true. Reliability API, claim c_06a3e6342890, method 0.1.0+28da144b. https://epistacklabs.com/claims/c_06a3e6342890
Claims without a verdict
Is a verdict here wrong, or do you speak for Daniela Amodei? 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.