Vinod Khosla

And that three-dimensional molecule, that big chunky protein does something structurally, physically. And so trying to understand the shape of a protein is really hard. I mean, we've used kind of x-ray imaging systems to try and identify it and tried to build models to identify how does that quote protein folding work?

And that three-dimensional molecule, that big chunky protein does something structurally, physically. And so trying to understand the shape of a protein is really hard. I mean, we've used kind of x-ray imaging systems to try and identify it and tried to build models to identify how does that quote protein folding work?

How do those amino acids collapse on each other to create that three-dimensional construct? And I don't know if you guys remember in the early 2000s, there was a Stanford folding at home distributed computing project. Do you guys remember this? Yeah.

How do those amino acids collapse on each other to create that three-dimensional construct? And I don't know if you guys remember in the early 2000s, there was a Stanford folding at home distributed computing project. Do you guys remember this? Yeah.

Exactly right, yeah. So it was like, it ran on the background of your computer, it used your CPU cycles when you weren't using your computer, and it tried to model protein folding. And so this has been a problem that folks have tried to tackle with compute for decades to figure out the 3D structure.

Exactly right, yeah. So it was like, it ran on the background of your computer, it used your CPU cycles when you weren't using your computer, and it tried to model protein folding. And so this has been a problem that folks have tried to tackle with compute for decades to figure out the 3D structure.

This is so important because if we can identify the 3D structure of proteins and we can predict them from the amino acid sequence, we can print out a sequence of amino acids to make a protein that does a specific thing for us.

This is so important because if we can identify the 3D structure of proteins and we can predict them from the amino acid sequence, we can print out a sequence of amino acids to make a protein that does a specific thing for us.

And that unlocks this ability for humans to create biomolecules that can do everything from binding cancer, to breaking apart pollutants and plastics, to creating entirely new molecules, to running, in some cases, like what David Baker did at University of Washington, he shared the Nobel Prize, creating micro motors, mini motors, from proteins that he designed on a computer.

And that unlocks this ability for humans to create biomolecules that can do everything from binding cancer, to breaking apart pollutants and plastics, to creating entirely new molecules, to running, in some cases, like what David Baker did at University of Washington, he shared the Nobel Prize, creating micro motors, mini motors, from proteins that he designed on a computer.

And so this becomes, I think this great, like big Holy grail in biochemistry and the alpha fold project at deep mind inside of Google solve this problem. And, and by the way, since, since then they've come out with alpha fold three, they've launched a drug discovery company called isomorphic labs, where they're basically predicting molecules that will do specific things for a target indication.

And so this becomes, I think this great, like big Holy grail in biochemistry and the alpha fold project at deep mind inside of Google solve this problem. And, and by the way, since, since then they've come out with alpha fold three, they've launched a drug discovery company called isomorphic labs, where they're basically predicting molecules that will do specific things for a target indication.

Uh, and then they use the alpha fold models to actually design and develop those molecules.

Uh, and then they use the alpha fold models to actually design and develop those molecules.

And there have been literally dozens of companies that have been started since DeepMind was published, and probably several billion dollars of capital that's gone into companies that are creating new drugs, creating new industrial biotech applications, using this protein modeling capability that was unleashed with DeepMind a number of years ago. So it really has transformed the industry.

And there have been literally dozens of companies that have been started since DeepMind was published, and probably several billion dollars of capital that's gone into companies that are creating new drugs, creating new industrial biotech applications, using this protein modeling capability that was unleashed with DeepMind a number of years ago. So it really has transformed the industry.

It'll be a couple years before we see it transforming the world, but it's an exciting kind of thing, yeah.

It'll be a couple years before we see it transforming the world, but it's an exciting kind of thing, yeah.

Anyway, congrats to the teams. Congrats to Demison Jumper. Yeah. I mean, it's just great. And David Baker at the Baker Lab in University of Washington.

Anyway, congrats to the teams. Congrats to Demison Jumper. Yeah. I mean, it's just great. And David Baker at the Baker Lab in University of Washington.