By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
No PriorsPowering AI-Driven Drug Discovery with President of Isomorphic Labs Max Jaderberg
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
Great conversations run long. Their best ideas don't have to.
uListen is free on iPhone, iPad, Android, Chrome, and Claude.
See howAt a glance
WHAT IT’S REALLY ABOUT
Isomorphic’s plan to move drug discovery from labs to computers
- Max Jaderberg described drug discovery as a historically empirical, trial-and-error process that struggles with biological complexity and has diminishing “low-hanging fruit.”
- He argued that the biggest cost driver is late-stage failure in clinical trials, where toxicity, off-target effects, and mistaken biological assumptions often only become evident in patients.
- Isomorphic Labs’ core aim is to bring drug discovery onto computers by building predictive models that can match experimental accuracy, so real-world experiments shift toward validation rather than iterative learning.
- He said Isomorphic combines top AI talent and life-science experts, develops its models in-house, and pursues both an internal drug pipeline and partnerships with pharma/biotech teams.
- He discussed rethinking traditional preclinical endpoints (including reliance on animal predictivity) and pointed to combining in-vitro measurements, organoids/organs-on-chips, and AI models as a direction for better human-relevant prediction.
IDEAS WORTH REMEMBERING
5 ideasDrug discovery is still fundamentally an empirical trial-and-error process.
He argued that discovery is still largely trial-and-error: you form hypotheses, synthesize molecules, and test iteratively in real assays, which is slow and noisy in a high-dimensional biological and molecular design space.
The clinic is where many drugs fail because problems aren’t visible earlier.
He said failures most often show up only once drugs reach patients, when unexpected toxicity, off-target effects, or flawed biological hypotheses become apparent, driving high average costs because so many programs fail late.
The goal is experimental-accuracy simulation so iteration happens on computers, not in labs.
He described Isomorphic’s “North Star” as moving drug discovery work onto computers by building predictive “world models” that reach experimental accuracy, so teams can do more iteration in silico and reduce real-world experimental turns.
Data needs for generalizable biology models often require generating new, purpose-built datasets.
He said their models are built in-house and powered by a comprehensive data strategy that includes generating proprietary data, because many existing repositories are biased by being passive byproducts of past drug-making rather than optimized for generalization.
Robotic labs help, but physics limits how fast real-world iteration can get.
He framed automation in wet labs as useful but ultimately bounded by physical constraints like reaction times and cell culture timelines, so it should complement—rather than replace—the push to reduce dependence on the real world via in-silico prediction.
WORDS WORTH SAVING
5 quotesAnd so the really big opportunity is to have much more of that understanding before you go into patients.
— Max Jaderberg
So, um, you know, if you, if you go into our like, uh, sort of our core AI team and, uh, like our North Star and the thing that we're really trying to achieve is how can we bring all of that drug discovery work, all of that drug design work, understanding of, you know, disease biology onto a computer?
— Max Jaderberg
Fundamentally, I wanna get out of the real world, right?
— Max Jaderberg
If you really care about generalization out of these models, there's maybe a different distribution of data you need.
— Max Jaderberg
I'd be really disappointed if we fast-forward, you know, even five years' time, if we don't have the base tools like the, like the predictive understanding of disease biology has not step change in, like, a massive way.
— Max Jaderberg
Ask ChatGPT about this episode
Get its key ideas, plus a summary of each chapter linked to that point in the video, right in your chat.
Add uListen AI to ChatGPT.
Install uListen AI in ChatGPT (opens in a new tab)Needs a ChatGPT account.Paste this prompt into a new chat.
Use uListen AI to give me this episode's key ideas and a chapter-by-chapter summary, with timestamps: https://www.youtube.com/watch?v=IiKf3BgBezE
SPEAKERS
Max Jaderberg
guestPresident of Isomorphic Labs and former DeepMind researcher focused on AI systems for drug discovery.
Elad Gil
hostInvestor and entrepreneur (co-founder/early CEO of Color Health) who co-hosts No Priors.
High quality AI-generated summary created from speaker-labeled transcript.
