YouTube · The Diary Of A CEO · 120:50 · Jul 22, 2026
The interview is built around Daniel Kokotajlo’s views and forecasts about AI risk. His background and the existence of his AI forecasting work check out, but several of the biggest claims are predictions or exaggerations, not established facts.
Claim-by-claim breakdown
Accurate (95% confidence): Daniel Kokotajlo previously worked at OpenAI and now leads the AI Futures Project. AI Futures Project’s official site says it is led by Daniel Kokotajlo, a former OpenAI governance researcher.
Accurate (84% confidence): OpenAI had a non-disparagement clause tied to employee equity, and the issue led to public backlash in 2024. WIRED reported that OpenAI faced criticism over non-disparagement agreements tied to equity clawbacks, and that Altman said the clause would be removed.
False (96% confidence): Anthropic’s revenue went from about $1 billion a year to about $60 billion in one year. Official Anthropic posts show much lower revenue figures than the transcript claims, and the timing does not match a 1B-to-60B jump in one year.
Accurate (98% confidence): AI Futures Project published a report called AI 2027 and a follow-up called AI 2040 plan A. Official AI Futures Project pages confirm both projects and connect them to Kokotajlo’s group.
Unsupported (93% confidence): The claim that AI has a 70% chance of causing human extinction is a factual forecast. No empirical source can verify a future extinction probability; at most, the speaker’s forecast can be quoted as his opinion.
Misleading (90% confidence): Anthropic is on track to be the entire economy by 2030. Official Anthropic revenue announcements show rapid growth, but they do not support the literal claim that the company will become the whole economy.