Gary Marcus argues Dario Amodei’s medical predictions overstate what AI can deliver within a decade

Gary Marcus argues Dario Amodei’s medical predictions overstate what AI can deliver within a decade

Gary Marcus argues that Anthropic CEO Dario Amodei’s prediction that cancer, Alzheimer’s disease, heart disease and mental illness will be “relics of the past” in less than a decade is an implausible overpromise. Marcus points to limits in biological data and the time required for drug development and clinical trials. He cites Eli Lilly CEO Dave Ricks, who says the biology map is “still largely a blank page,” randomized human proof is “the long pole in the tent,” and progress is likely to be steady and incremental. Marcus presents this criticism as coming from both AI skeptics and researchers who are optimistic about AI’s potential in health. The distinction matters in his account: optimism about AI improving health is not the same as confidence that it will cure several major diseases on Amodei’s timeline. Marcus also connects the prediction to Anthropic’s IPO, arguing that media outlets should scrutinize such claims and that the timing makes them especially concerning. The article ends by introducing a software-development example that is not included in the supplied text.

Publication: fcc074e1-2b30-4ed9-8a46-c8c295f138a5

attribution — pass: The reading correctly attributes the argument to Gary Marcus. It attributes the 'relics of the past' prediction to Amodei, as quoted in the embedded post. It correctly attributes the 'blank page' and 'long pole in the tent' remarks to Eli Lilly CEO Dave Ricks. Ricks says 'our sector'; rendering that as health is a reasonable reading.

Here’s the Chairman annd CEO of Eli Lilly, Dave Ricks, critiquing Dario and giving a far more realistic view:

summary — uncertain: The summary weakens Marcus's central force. The title and body repeatedly compare Amodei to Elizabeth Holmes, and Marcus says the claims are 'so egregious' they should be investigated for securities fraud. The summary only says media should scrutinize the claims and that the timing is 'especially concerning'. The Holmes comparison, the core of the headline, is missing. Smallest repair: add that Marcus likens the claims to Elizabeth Holmes, the Theranos founder, and says they should be investigated for securities fraud, keeping both as his assertions. The rest is accurate and accessible.

which is so egregious it should be investigated for securities fraud — the media is failing the American public.

date — pass: The originalPublished value, 2026-11-12, matches the article byline date and the feed entry. The embedded posts from Nov 10–11 predate it, consistent with the article.

Gary Marcus Nov 12, 2026

support — pass: Each claim is grounded in the source. Claim 2's excerpt, 'The data exists for', is very short and comes from Rich Law's quoted post, not Marcus's own words. However, Marcus frames it with 'we don't have the data' and argues that clinical trials cannot be expedited, so the claim is supported. A fuller excerpt would help. The question is open, not leading. The limitations accurately note that the source gives no independent assessment.

As several researchers noted, we simply can’t expedite clinical trials — which are absolutely necessary — to the degree that would be required

media — pass: The selected media is the verified direct link to the original Substack article. It is the most direct way to access the source.