Marc Tessier-Lavigne

speaker

407 appearances 1 recordings 1 series first heard Mar 2026 last heard 12 Mar

Marc Tessier-Lavigne’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months

Recordings per month over the last 12 months — 1 in all, peaking in Mar 2026 with 1.

Appearances

newest first · ▶ plays the moment

Vanguards of Healthcare by Bloomberg Intelligence · Xaira’s War on Trial-and-Error Drug Development · 12 Mar 2026 podcast

For example, not just which mRNAs are there and how they interact, but which proteins are actually present and other features of the cell.

So I think the field as a whole will be able to make progress by creating appropriate causal models that will become more and more and more accurate, richer and richer in data to provide better and better answers.

But we believe that in the near term, we'll be able to get some answers from these initial causal models of the cell.

Yeah, absolutely.

That's a great question and exactly the kind of path that we're on.

So the story of AI is you train on massive amounts of data that you can, I was going to say that are easily available.

None of this is easily available, but there's some that you can garner in very large amounts in relatively straightforward manner as we have been with these perturbation experiments.

Those experiments done in cells, and especially experiments done in cells in a petri dish, are removed from the ultimate goal of understanding what's happening in a human being.

And we have various models of increasing degrees of complexity.

You go from single cells to organoids, which, of course, is another technological advance in the past century.

decade, or even into animal models where you can do experiments, perturbation experiments in the living organism, in a mouse, for example.

And finally, to the human, where it may not be possible to do that directly in the human, but you may be able to use surrogates, certainly use tissues from patients.

With AI, what you do is you first build a model with massive amounts of data that are readily available.

That's the pre-training.

You can then post-train with data that might be closer to what you're most interested in that are harder to come by.

And you can do further fine-tuning with data that are very hard to come by, like disease tissues from patients.

And the extraordinary thing about AI is to see the transfer learning that occurs between those different levels.

Let me give an analogy with ChatGPT.

The ChatGPT trained on, you know, all the world's information.

But

Showing 221–240 of 407 · page 12 of 21