Isaac Kohane
@zakkohane
Computer science + Biomedicine = Medicine as a Knowledge Processing Discipline @harvardMED @HarvardDBMI Prof/Dept Chair http://zaklab.org MD/PhD @NEJM AI EIC
Paradox also highlighted recently by @pmarca on @a16z pcast, as I discussed in @timmermanreport -> timmermanreport.com/2025/02/the-fu… cc @zakkohane @sapinker @emollick @vijaypande
Observed paradox in AI in medicine @NEJM_AI ai.nejm.org/doi/full/10.10… but I suspect as fast as you move the goal posts, AI will be there soon enough. @emollick
Observed paradox in AI in medicine @NEJM_AI ai.nejm.org/doi/full/10.10… but I suspect as fast as you move the goal posts, AI will be there soon enough. @emollick
A crucial point about human abilities, which I emphasize in my intro psych lectures: In AI, the "hard problems" (in this case, acrobatics) are easy, and the "easy problems" (doing useful things with objects) are hard.
Why did @MIT_CSAIL 6.001 switch to Python from Scheme? This exerpt from the International Lisp conference of 2009 cuts deep wingolog.org/archives/2009/… via @TheHackersNews The "debate" had an interlude, in which Costanza asked Sussman why MIT had switched away from Scheme for…
Medical students re-imagining what their role is in medicine thoughtfully framed by @morgancheatham at ~14 minutes into the podcast 👇
In Episode 9 of "The AI Revolution in Medicine, Revisited," resident physician and venture capitalist Morgan Cheatham and med student Daniel Chen explore generative AI's impact on medical training, from classroom learning to clinical practice. msft.it/6011s1Qrd
I think this might be my favorite episode of this podcast series. Talking to current med students about their thoughts and actions on AI in medicine is enlightening. It's impressive how they are really leaning in and learning together.
In Episode 9 of "The AI Revolution in Medicine, Revisited," resident physician and venture capitalist Morgan Cheatham and med student Daniel Chen explore generative AI's impact on medical training, from classroom learning to clinical practice. msft.it/6011s1Qrd
Which way to the future: Everyone has their own favorite AI aligned to their personal (healthcare) interests and therefore stoking an arms race to the most persuasive, legally-expert AI? Or one AI (informed by which powers/interests?) to adjudicate for them all? HT to…
𝐀𝐈 𝐢𝐬 𝐡𝐞𝐥𝐩𝐢𝐧𝐠 𝐩𝐚𝐭𝐢𝐞𝐧𝐭𝐬 𝐟𝐢𝐠𝐡𝐭 𝐢𝐧𝐬𝐮𝐫𝐚𝐧𝐜𝐞 𝐜𝐨𝐦𝐩𝐚𝐧𝐲 𝐝𝐞𝐧𝐢𝐚𝐥𝐬 Insurers participating in the Affordable Care Act denied 19% of in-network claims in 2023. AI empowers patients to fight back ⤵️ nbcnews.com/news/us-news/a…
To @TomKnightSynBio's point, there was a lot of (premature) excitement and many of us were all in.
You, and seemingly the world, forget that AI was a very hot topic in the 1980s. We were in a technology battle with Japan's Fifth Generation computer effort, and expert systems, for most everything, were the rage.
You, and seemingly the world, forget that AI was a very hot topic in the 1980s. We were in a technology battle with Japan's Fifth Generation computer effort, and expert systems, for most everything, were the rage.
Wait… what?! 🤯 The first domain name ever was Symbolics.com. It belonged to a company making Lisp machines... purpose-built for AI before AI was cool. They powered early breakthroughs in robotics, language, and logic. Now AI is everywhere. The internet literally…
Example of AI breaking artificial siloes of medical expertise. @morgancheatham made this point eloquently in his recent @NEJM AI Grand Rounds podcast.
Congrats @David_Ouyang on what will be transformational work at the intersection of CKM health, evaluating #opportunistic screening & #ImplementationScience to improve population health!! @American_Heart @HeartDocSadiya @KPDOR newsroom.heart.org/news/researche…
SPIRO-AID trial: AI reads lung tests better than doctors. Even better than doctors using AI? @NEJM_AI ai.nejm.org/doi/full/10.10…
Incredibly honored and grateful to receive the Overton Prize at #ISMBECCB2025♥️ Many thanks to ISCB and my amazing students, collaborators and mentors!🙏
Back to Susan Calvin zaklab.org/blog/embrace-y… HT @arjunmanrai
We're launching an "AI psychiatry" team as part of interpretability efforts at Anthropic! We'll be researching phenomena like model personas, motivations, and situational awareness, and how they lead to spooky/unhinged behaviors. We're hiring - join us! job-boards.greenhouse.io/anthropic/jobs…
We let the web rot away well before LLMs This chart shows the percentage of links from all New York Times articles that still work. Over 60% of older links are now broken. And consider that social media posts are even more ephemeral Likely only LLMs will “remember” that content
Nomen est omen. He was annointed early (see nomenomen.us)
it really is destiny.
"Baseline CAD (wait, isn't that the outcome?)" is a little concerning if it indeed contributes to the prediction of CAD.
AI-assisted primary care in Kenya racks up impressive reduction in diagnostic and therapeutic errors. When do we fix primary care in Boston and Palo Alto?
📣 Excited to share our real-world study of an LLM clinical copilot, a collab between @OpenAI and @PendaHealth. Across 39,849 live patient visits, clinicians with AI had a 16% relative reduction in diagnostic errors and a 13% reduction in treatment errors vs. those without. 🧵
Generate top viewership for unusual slices of RealityTV™ @nvidia @xai @DellTech @MrBeast given challenging elimination rounds for human+AI teams and a final round goal that would have positive societal impact. Variant: teams compete by sending their creations into SOTA world…
nvidia could do the most viral ai competition in history: start with 10,000 researchers and give each a free gpu to work on a public leaderboard but do rounds of elimination where the winners take the remaining hardware. the final winner gets all the gpus for a year.
nvidia could do the most viral ai competition in history: start with 10,000 researchers and give each a free gpu to work on a public leaderboard but do rounds of elimination where the winners take the remaining hardware. the final winner gets all the gpus for a year.