Will AI make major diseases 'relics' within a decade?
Dario Amodei told investors that cancer, Alzheimer's, heart disease and mental illness will be 'relics' in under a decade. Anthropic's published results are early lab work, and a drug that reaches approval takes 10.5 years from its first human dose on average.
I don't believe the claim that cancer, Alzheimer's, heart disease and mental illness will be relics within a decade, and I read it as a sales pitch. AI is speeding up the early part of drug discovery. It cannot skip clinical trials, and the trials set the clock. What would change my mind is a drug that Claude helped design clearing a late-stage human trial in one of those diseases, which Anthropic has not reported.
On 10 November 2026 Anthropic launched the roadshow for its IPO, the round of investor presentations that comes before shares are priced1. In the roadshow video, chief executive Dario Amodei says these diseases "will be relics of the past in, I think, less than a decade", and adds "I think if we do this right, we can create an age of miracles". TechSpot reported the line on 12 November, and a transcript of the video's captions places it at 40 minutes 37 seconds23. The launch announcement says nothing about disease1. The written pitch on Anthropic's IPO site says AI could bring "cures for major diseases within reach", with no date attached4. Only the video gives a date. CNBC reported on 10 November, citing unnamed sources, that Anthropic planned to price on 18 November5, and the IPO site now gives the same date4. That puts the claim in an investor video eight days before pricing.
The pushback came fast. On 11 November Dave Ricks, chief executive of Eli Lilly, replied that the claim would be amazing but is "not likely", because "the map of biology is still largely a blank page and randomized human proof is the long pole in the tent". He said he is excited about AI in the sector and expects progress to be "steady and incremental"6. Lilly is an Anthropic customer, and its chief information officer is quoted on the IPO site4. On 10 November Eric Topol, a physician-scientist, wrote that "curing all these diseases in the next 10 years is impossible", and added that he is "very optimistic about the impact of AI for improving health"7. Karl Lauterbach, Germany's former health minister, wrote in German on 11 November that these diseases can benefit from AI but a cure cannot possibly arrive in just 10 years (my translation)8.
On 12 November Gary Marcus, a long-time AI sceptic, called Amodei the "heir to Elizabeth Holmes" and said the claims should worry you about the IPO9. I would not go that far. Amodei said "I think", and the written pitch carries no timeline. On Hacker News a recurring objection was the loop between the lab and the patient. One commenter put it that medicine is slow for AI for the same reason it is slow for people, because the test-and-result loop is slow10.
On 27 October 2026 Anthropic published results from six partners and its own programmes, and some of it is impressive. At Generate Biomedicines, six scientists using Claude Code proposed a design for part of a CAR T-cell therapy, a cancer treatment built from a patient's immune cells. It was about 70% shorter than the standard version and let immune cells make roughly 10 times as many receptors in lab tests11. Divergence Labs had Claude generate about 35,000 antibody designs in about a month for a food-allergy drug, and found three that bound IgE and 20 that dislodged it11. BigHat Biosciences made and tested about 1,010 variants in about 10 weeks, and the best bound more than 10 times more tightly11. At Enveda, review of manufacturing paperwork fell from over 25 hours to under three11. Anthropic's own muscular dystrophy programme, which came with its acquisition of Mithril Medicine, went from an idea to "promising results in mice" in under 18 months11. That is faster design and faster paperwork, and I take it seriously.
It is also early. Anthropic says the work is "mostly preclinical", meaning for the most part not yet tested in humans, and every figure above is reported by the company or its partners11. If the Mithril candidates keep holding up, Anthropic expects human trials "as soon as late 2028"11.
Human trials are where the time goes. BIO, Informa Pharma Intelligence and QLS Advisors tracked clinical programmes over 2011 to 2020. Only 7.9% of drugs entering Phase I, the first safety trials in people, reached FDA approval. The ones that did took 10.5 years on average12. For the groups closest to Amodei's list the odds were lower.
Share of drugs entering Phase I that reached FDA approval, by disease group, 2011 to 2020 (BIO, Informa Pharma Intelligence and QLS Advisors, February 2021). Neurology covers more than Alzheimer's and the data stops at 2020
The arithmetic here is mine. A drug that entered Phase I this month would, on the average timeline, be approved around May 2037, after Amodei's decade ends. The Mithril candidates, if human trials start in late 2028 and they run to average, would land around 2039. That programme targets a rare disease outside Amodei's list, and an approved drug is still a long way from a disease becoming a relic.
Anthropic has only just started on the trial stage. Its work with Bristol Myers Squibb on clinical trial planning is due to expand in 202711.
I expect AI to make drug discovery faster, and I would be happy to be wrong about the decade. Until a drug Claude helped design reads out a Phase III result in one of those four diseases, I will treat the decade as a number from a roadshow.
Nothing here is financial, legal or investment advice. Every figure is published by its named source and carries the date it was published; figures may since have changed.
References
-
Anthropic, Anthropic launches roadshow for proposed initial public offering ↑
-
TechSpot, Anthropic's Dario Amodei uses IPO pitch to promise AI will eliminate cancer and Alzheimer's within a decade ↑
-
Blumify, Anthropic IPO - Dario Amodei (transcript of the roadshow video's captions) ↑
-
Anthropic, Anthropic's Initial Public Offering ↑
-
CNBC, Anthropic sets IPO price range of $70 to $78, looks to start trading Nov. 19, sources say ↑
-
@DaveARicks on X, Reply to Meg Tirrell on Amodei's "relics" claim (11 Nov 2026) ↑
-
@EricTopol on X, Reply to NIK on the 10-year disease claim (10 Nov 2026) ↑
-
@Karl_Lauterbach on X, Post on Amodei and curing disease in 10 years (11 Nov 2026) ↑
-
@GaryMarcus on X, Post on Amodei, Elizabeth Holmes and the IPO (12 Nov 2026) ↑
-
Hacker News, Eli Lilly CEO pushes back on Dario Amodei's medical claims ↑
-
Anthropic, Early progress on accelerating drug development with Claude ↑
-
BIO, Informa Pharma Intelligence and QLS Advisors, Clinical Development Success Rates and Contributing Factors 2011–2020 ↑