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?
76 69 82 -
Predictions
Statements about events outside the person's control, checked once they fell due: did they come true?
56 42 70 -
Promises
Commitments to the person's or organisation's own future action, with something to deliver: was it delivered?
Insufficient record: 4 promises resolved, 5 needed for a score.
All types combined: 72 (likely 66 to 78, high confidence), from 93 claims: 76 factual claims, 17 predictions. It mixes kinds of claim in different proportions for different people, so compare people kind by kind.
Dario Amodei's claims
28 claims
“我们当前的等级是 A.S.L. 2。”
Dario Amodei, Anthropic CEO, interviewee 12 April 2024 Transcript: Ezra Klein Interviews Dario Amodei (translation) (baoyu.io)
Anthropic's models are currently at AI Safety Level ASL-2.
Resolved true The record shows it happened or was so.
Said for an organisation they lead or work for; it counts on both records.
What would count as true or false
True: as of 2024-04-12, Anthropic's official materials (Responsible Scaling Policy, model cards, system cards for Claude 3 models) show its deployed models were assessed at, and covered by, ASL-2 standards. Mostly true: models were under ASL-2 with a minor caveat (for example some precautionary ASL-3 preparation under way) that does not change the classification. Mixed: not used for this single-fact claim. Mostly false: Anthropic's records show the classification at that date was ambiguous or that some current models were already treated under ASL-3. False: Anthropic records show its models were at ASL-3 or higher (or ASL-1) as of that date.
Evidence
- Introducing Anthropic's Responsible Scaling Policy Anthropic · 19 September 2023 Own record
“Current LLMs, including Claude, appear to be ASL-2.”
- Responsible Scaling Policy | AI Wiki AI Wiki (aiwiki.ai) · 30 April 2026
“Classified ASL-2 in the Claude 3 model card after Anthropic's evaluations showed no indicators of catastrophic risk.”
Reasoning
Anthropic's Responsible Scaling Policy announcement (E1) says that current LLMs, including Claude, appear to be ASL-2. An independent reference (E2) reports that the Claude 3 model card classified those models as ASL-2 because Anthropic's evaluations showed no signs of catastrophic risk. Neither source shows any deployed model being treated under ASL-3 at the time. Together they indicate that Anthropic's models were classified as ASL-2 around April 2024, although E1 says only that they 'appear to be' ASL-2 and E2 is a secondary summary of the Claude 3 model card.
Cite this claim
Dario Amodei, “Anthropic's models are currently at AI Safety Level ASL-2.” (said 12 April 2024). Verdict: resolved true. Reliability API, claim c_d672b81f4327, method 0.1.0+c876030c. https://epistacklabs.com/claims/c_d672b81f4327
Claims without a verdict
- Unresolvable 157
- Not yet due 40
- Own records only 43
- Not researched 134
- Not checkable 864
Is a verdict here wrong, or do you speak for Dario 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.