Skip to content

By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.

Silicon Valley GirlSilicon Valley Girl

AI’s Next Trillion-Dollar Market Isn’t Software | Max Jaderberg, President of Isomorphic Labs

AI can generate thousands of ideas. I asked Max Jaderberg how to tell which ones are worth pursuing.

Max JaderbergguestMarina Mogilkohost
Nov 10, 202643mWatch on YouTube

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 how

At a glance

WHAT IT’S REALLY ABOUT

How Isomorphic Labs uses AI agents to redesign drug discovery

  1. Max Jaderberg explains how AlphaFold built confidence that AI can generalize in biology and chemistry, enabling Isomorphic Labs’ mission to redesign how medicines are made.
  2. He describes drug development as a long, failure-prone process with many interlocking scientific challenges, requiring multiple “AlphaFold-like breakthroughs” rather than a single model.
  3. He outlines a drug design pipeline where generative/design agents propose many molecules and complementary “virtual lab” models filter them down before expensive real-world synthesis and testing.
  4. He details how Isomorphic pushes commercialization early via partnerships and by continuously translating internal benchmarks into measurable improvements on live drug discovery programs.
  5. He discusses practical organizational and talent approaches—mission trees, quantitative leaderboards, interdisciplinary hiring, and eval-driven agent development—while emphasizing that human attention and people leadership remain central.

IDEAS WORTH REMEMBERING

5 ideas

AlphaFold mattered less as a drug tool and more as evidence AI can generalize in life sciences.

He describes AlphaFold as a “proof point” that AI models can generalize in biology and chemistry, which gave him conviction that the field is ready to build toward an AI-driven drug design engine.

Drug creation is a multi-problem system, not a single-model challenge.

He emphasizes that making medicines requires many “really tricky problems” to come together (target selection, molecule design, delivery to the right body location, safety/toxicity, etc.), which is why the process takes about 10 years and has a very high failure rate.

The core workflow is generate broadly, simulate to filter, then validate in the wet lab.

He frames Isomorphic’s work as combining models that propose many candidate designs with “virtual lab”/simulation models that predict outcomes, so the team can narrow from massive hypothesis spaces to a small set of molecules to synthesize and test.

Commercial readiness is about benchmark-to-reality transfer, not absolute accuracy.

He says their decision rule for “ready for real work” isn’t perfection; because the industry baseline is trial-and-error, a meaningful predictive improvement over the status quo can create value, but only if benchmark gains translate into real program outcomes.

Humans increasingly specify goals and constraints; agents do the search and iteration.

He explains that researchers’ roles are shifting toward defining objectives and design spaces while agents explore the options—both for AI research (agents running experiments) and for chemistry (agents generating molecule variants).

WORDS WORTH SAVING

5 quotes

AlphaFold isn't the solution to making medicines, to solving diseases. It's really that proof of concept that a solution could exist.

— Max Jaderberg

The solution from a AI perspective and a model perspective is not, is not actually just one model. It's like a whole collection of AlphaFold-level breakthroughs.

— Max Jaderberg

Yeah, our mission is crazy. You know, solve all disease. It's, it's, it's pretty audacious. Um, we say that with a straight face. Like, we really, like, feel and mean that mission.

— Max Jaderberg

I'm a deep learning maximalist. I think we have the algorithms that can solve almost any problem. We have the compute. Now we have just an abundance of compute.

— Max Jaderberg

It's at the point in time now where instead of having to screen, experimentally go through hundreds of thousands of different designs, we can make 20 or 50.

— Max Jaderberg

Read the available transcript
AlphaFold as proof of generalization in biologyWhy drug development is slow and failure-proneFoundational biology models vs drug design modelsGenerative design agents and virtual lab/world modelsBenchmarks, evals, and benchmark-to-reality translationPartnerships vs internal drug programs and IPInterdisciplinary hiring and learning ability

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.

  1. Add uListen AI to ChatGPT.

  2. 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=yIFhTVMOH98

SPEAKERS

  • Max Jaderberg

    guest

    President of Isomorphic Labs and AI researcher (ex-DeepMind), focused on applying deep learning to drug discovery.

  • Marina Mogilko

    host

    Host of Silicon Valley Girl, interviewing founders and tech leaders about startups, AI, and innovation.

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

Read along on your phone, search it beside the video, or ask your favorite AI app what was said.