The creator explains how AI, specifically DeepMind's AlphaFold, could cure all diseases within a decade by solving the protein folding problem. He tells the origin story of DeepMind's CEO, Demis Hassabis, from child prodigy to Nobel laureate, highlighting his mission-driven approach and how his work is now tackling the next phase of drug discovery.
Post caption
I sat down with Demis Hassabis — Nobel laureate, DeepMind CEO, and arguably the most important scientist alive — who famously said that the end of disease could “happen in the next decade or so.”
Which sounds impossible, but when you understand the way he thinks about it, the question shifts from if to when.
In this video, I break down why so many people in the scientific community believe that the “end of disease” is very plausible.
If you want to listen to my conversation with Demis on our new show @frontier, comment “CURE” and I’ll send over the link.
Note: Demis shared the 2024 Chemistry prize with John Jumper and David Baker.
Transcript
AI will cure every disease within the next ten years. Cancer, Alzheimer's, all of it. At best, that's what Dennis Hassabis believes. And as the Nobel Prize winning CEO of DeepMind, you might be onto something. See, everything happening in your body runs on proteins, from your muscles contracting to your immune system killing a virus. If you understand proteins, you understand disease. The problem is that it took scientists sixty years to map just one percent of them. But Demis built an AI called AlphaFold that mapped the other 99% in about a year. 200,000,000 structures that were supposed to take centuries. Then he gave away the entire database to the world for free, which won him the Nobel Prize in 2024. But the mission started long before that. Demas taught himself to code at eight years old, became the second best 14 chess player on the planet, got into Cambridge at 16, and by 2010, he founded DeepMind, back when AGI was still science fiction in most of the world. When Google bought DeepMind for $650,000,000, he had one nonnegotiable. DeepMind continued doing fundamental science to push humanity forward because the money was never the point. The mission good. And now that mission is aimed at the biggest prize of all because mapping proteins is only step one. See, even after you find a drug, it takes ten years of clinical trials and a billion dollars to prove that it works. Most of them fail because nobody can predict how a molecule will behave inside the human body until they test it. That's the problem Demis is going after now. If an AI can model exactly how a drug latches onto its target before anyone runs a trial, then the molecules that make it to testing are the ones most likely to actually worse. A search that took years collapse into weeks. No when I sat across from Demis, he told me he barely sleeps and can't remember his last holiday because to him, the mission is worth everything it costs. Now whether he's right is yet to be seen, but everybody said that mapping every protein was impossible. He did it in a year, and now he's aiming at every disease on Earth. If you wanna listen to my full conversation with Demis, comment cure, and I'll send you a link. Catch you in the next one. Peace.
