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
49 claims
“Which I think had real commercial consequences, but set the culture of the company”
Dario Amodei, Anthropic CEO 10 March 2025 CEO Speaker Series with Dario Amodei of Anthropic (Council on Foreign Relations)
Amodei said he thinks delaying Claude's release had real commercial consequences for Anthropic.
Mostly true True in substance, wrong in a detail.
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 Anthropic held back the release of an early Claude model (in 2022, before ChatGPT) and that this had real commercial consequences, i.e. lost first-mover position, market share, revenue or attention to competitors. True: credible accounts (reporting, histories, statements from investors or former staff) confirm the delay happened and describe concrete commercial cost, such as OpenAI's ChatGPT capturing the market first. Mostly true: the delay is confirmed and commercial cost is plausibly shown, but evidence of the size of the cost is thin or mainly inferred. Mixed: the delay is confirmed, but evidence on whether it cost Anthropic commercially is balanced or contested. Mostly false: the delay is confirmed, but evidence suggests little or no commercial effect, or the delay was mainly for other reasons such as the model not being ready. False: evidence shows no such deliberate delay took place.
Evidence
- Deep Dive: $ANT The Weng Lab (Substack) · 27 October 2026
“Months before ChatGPT existed, Anthropic had a working chatbot. Staff and friends used it inside Slack. People liked it.”
- Deep Dive: $ANT The Weng Lab (Substack) · 27 October 2026
““Our general feeling was that it would cause too much acceleration,” co-founder Ben Mann said later.”
- Deep Dive: $ANT The Weng Lab (Substack) · 27 October 2026
“In November 2022, OpenAI launched ChatGPT, and it became the AI product the whole world knew.”
- Deep Dive: $ANT The Weng Lab (Substack) · 27 October 2026
“Claude finally launched in March 2023. That year, Anthropic made $41 million. OpenAI had a head start, a household name and Microsoft’s money.”
Reasoning
An independent newsletter reports that Anthropic had a working chatbot that staff used internally months before ChatGPT existed [E1]. It quotes co-founder Ben Mann as saying it was not released because it "would cause too much acceleration" [E2], which supports a deliberate delay. It also reports that ChatGPT launched in November 2022 and became the product the world knew, that Claude launched in March 2023, and that by then OpenAI had a head start and a household name [E3, E4]. That makes a commercial cost plausible, but the source does not measure any lost revenue or market share. All of this comes from one secondary source, so the evidence on how large the cost was is thin.
Cite this claim
Dario Amodei, “Amodei said he thinks delaying Claude's release had real commercial consequences for Anthropic.” (said 10 March 2025). Verdict: mostly true. Reliability API, claim c_85f421264e95, method 0.1.0+c876030c. https://epistacklabs.com/claims/c_85f421264e95
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.