Skip to content
Home » Transcript: 160 Years of Aging Research w/ Dr. David Sinclair @ Impact Theory

Transcript: 160 Years of Aging Research w/ Dr. David Sinclair @ Impact Theory

Editor’s Notes: In this fascinating episode, Tom Bilyeu sits down with Harvard geneticist Dr. David Sinclair to discuss how AI is revolutionizing longevity research by compressing 160 years of work into just a few years. Dr. Sinclair explains how his lab used AI to screen billions of molecules to find chemicals capable of reversing the aging process, potentially turning expensive gene therapies into affordable daily treatments. The conversation delves into the “Information Theory of Aging,” exploring how cells can be reprogrammed to regain their youthful function and the exciting human trials on the horizon for 2026. (April 16, 2026)

TRANSCRIPT:

AI’s Role in Accelerating Aging Research

TOM BILYEU: You are at the intersection of my absolute fascination with health, which is where it’s intersecting right now with AI. I’ve heard you say that AI is making things possible in human longevity that previously weren’t. So what specifically has AI put on the table that wasn’t possible before?

DR. DAVID SINCLAIR: Well, the big thing is the speed that we can do things. We currently have technology that can reverse aging in animals, and we’ll find out this year if it works in people. But it’s an expensive technology. It uses genes and we have to introduce genes into the body or the eye in this case. That’s potentially hundreds of thousands of dollars to do that.

So what we wanted to do in my lab was democratize this technology. So how do you do that? Well, AI is helping. We’ve now screened probably about 8 billion virtual chemicals for one that will reverse aging so that instead of introducing genes, which is expensive, we could take a pill or rub it on our hair or our skin.

And I asked one of the AI sites, “How long do you think this would have taken in a normal world, pre-AI?” And it estimated it would have taken about 160 years for my team to have finished that experiment, and the cost would have been in the many billions of dollars.

TOM BILYEU: Why is that? Is AI just crunching numbers, pattern recognition? What is it that makes AI able to shorten the timeline?

DR. DAVID SINCLAIR: Yeah, well, the big one was we need to thank Demis Hassabis and his team, of course, for elucidating the structure of all of the proteins in the body. We didn’t have that until about 4 or 5 years ago. And now that we have those structures, those proteins, we can virtually dock billions of molecules into each of those proteins and find ones that modulate those proteins.

TOM BILYEU: Is this based on shape or mostly—

DR. DAVID SINCLAIR: —and charge. So we know the behavior of atoms and small molecules, and now we know proteins, and proteins are vibrating, so it’s a little complicated.

Understanding What AI Actually Does in Biology

TOM BILYEU: You say “we know.” So one of the things I want to know about AI is, is AI getting to the point where it understands the fundamental rules that govern biology, or is it just learning all of the patterns in the literature?

DR. DAVID SINCLAIR: Oh, it’s more than in the literature. It’s understanding the patterns in biology and how to extrapolate from atoms to molecules to proteins. That’s a big jump. We couldn’t do that more than 5 years ago. And now AI is really using that.

I mean, it’s partly AI, it’s partly brute force, just mathematics. But on top of that, you can add now intelligent agents that can take the results from those screens, as we call them. And we typically get hundreds of thousands of hits, as they’re called.

TOM BILYEU: And the hit is this shape matches this shape and allows the chemical reaction to transpire.

DR. DAVID SINCLAIR: Exactly. In this case, what we’re doing is virtually impossible — at least 5 years ago it was impossible. And that is that we’re trying to find one chemical that does the work of three that we currently have. So when we reverse aging in a mouse, we give it a cocktail down its throat of three chemicals.

And in drug development, finding one better chemical can cost hundreds of millions of dollars and years of work. And I’m asking my team and our collaborators, “Find one molecule that does what those three do and even better.” And then once you’ve found those hundreds of thousands, which ones are most likely to work? Because in the lab, it’s not that easy, and it’s very expensive to order and synthesize thousands, and especially hundreds of thousands of molecules.

TOM BILYEU: Order meaning online? Yes, I need to get some of these molecules in the lab so I can get the mouse to eat it.

DR. DAVID SINCLAIR: Well, we don’t test it on a mouse initially. That would be prohibitively expensive.

TOM BILYEU: What do you test it on?

DR. DAVID SINCLAIR: Cells. And that’s also where AI comes in. We’ve got an AI system we developed. It’s machine learning with a layer of AI that can look at cells from humans that we grow in the lab and we paint them so they’ve got colors and so we can see different shapes and things that are happening inside the cells live, or we kill them and stain them.

And then we use visualization to say, “Is that cell from a 92-year-old going back to look more like the cells from a 20-year-old?” And we’ve been working for about 3.5 to 4 years on that.

TOM BILYEU: Did you guys have to train your own model?

DR. DAVID SINCLAIR: Yes, we absolutely did. So we got the cells from these people who are young and old, and we trained the model and it took a long time to get that right.

TOM BILYEU: And it was just like, “This cell young, this cell old, this cell young, this cell old.” And then it gets to the point where you don’t have to tell it anything. It just looks at it, it knows the patterns, and it says, “Oh, that’s an old cell, young cell, whatever.”

DR.