Twitxy Saved

Max Jaderberg

Verified account This account is verified. Learn more

@maxjaderberg

President @IsomorphicLabs: solving disease with AI. Prev: Deepmind, Vision Factory (acq. Google), Oxford VGG.
London, England scholar.google.co.uk/citations?user… Joined April 2009
1.3K Following 21K Followers 549 Posts

Latest posts 5

Pinned post

Max Jaderberg Verified account @maxjaderberg
Breakthroughs in frontier science rarely come from a sudden stroke of luck. Every major leap in AI drug design is the product of years of method development, rigorous engineering, and learning from what came before.

Witnessing the rapid advances in deep reinforcement learning and frontier AI capabilities led to a deep conviction: we could apply these same computational principles to other complex fields, like the engineering of new medicines.

That thinking led to the formation of Isomorphic Labs in 2021. Now, five years on, we have compelling evidence that our AI can unlock therapeutics to a wide range of disease classes. We are confident this technology will fundamentally impact how new medicines are created, from early discovery all the way through development.

In this article, I share the massive opportunity we have to transform human health with AI, and offer a look at the real, measurable progress we’re already making today.
isomorphiclabs.com/articles/build…
27 replies 77 reposts 603 likes 117K views View on X
Max Jaderberg Verified account @maxjaderberg
To unlock the full impact of AI on human health we need to build both the technology and the biological dataset it will run on. We’ve been developing our AI, our drug design engine (IsoDDE), for the last 5 years. Today, by joining the Virtual Biology Initiative as a founding member, we will be part of a large effort to generate the AI-ready data that will power predictive biology models.
Isomorphic Labs Verified account @IsomorphicLabs Oct 7
Today, Isomorphic Labs joins the Virtual Biology Initiative (VBI) as a founding member.
We are helping build an open resource for the global research community to accelerate how we understand and treat disease.
Read more here: isomorphiclabs.com/articles/isomo…
View quoted post on X
17 replies 29 reposts 221 likes 9.8K views View on X
Max Jaderberg Verified account @maxjaderberg
Here’s a visualisation I built of the generative design agent we use every day at Iso. IsoDDE takes the goals of our drug designers and works autonomously to produce a set of pareto best molecules. Throughout the core generative design and optimisation process you can see a visualisation of the agent’s attention (heatmap) as it balances exploration-exploitation across chemical space (here just umap visualised) to find the best molecules (greener = better)
I’ve pulled out only one pareto-best molecule to exemplify this process, but I’ll showcase in a later video how we do this at scale (1/5)
0:50
10 replies 10 reposts 130 likes 13K views View on X
Max Jaderberg Verified account @maxjaderberg
Interesting to see a benchmark for discovery that seems to have a meaningful axis of difficulty (at least wrt frontier models). And given that humans can solve first try.
James Whittington Verified account @jcrwhittington Aug 12
We find that
1) Opus 5 & Fable marked a real increase in ability on these games, but they still struggle with the hardest tiers of games (Tiers 6 and 7)
2) Agentic harnesses like Prime don’t offer any benefit over a basic (non coding!) harness
3) There’s a large gap between closed-source (Opus 5) and open-source models (Kimi K3) and another big gap to smaller open-source models (Qwen 3.6 27B)

Finally, rule discovery really is the bottleneck in these games: when we give the ground-truth rules in plain language, even the previously weakest models score near perfect.

(4/5)
View quoted post on X