Why Computer Science Subsumed Biotech

**Vijay Pande** (0:00)
The first half of any big change like this is that the technology is built full stack by the first company. Could Google have sold their search engine software to IBM and made money? Like, probably not. That's something that they had to build this out full stack. Google even built their own data centers, their own servers. They went really full stack in a sense. And I think what we're going to see on the healthcare side is something similar, that the early AI companies will actually also be the ones doing care delivery. Because if you're a full stack company, you could see the benefits of AI. You don't have to spend your time convincing the other people of it.

**Derrick Harris** (0:39)
Welcome again to the a16z AI podcast. I'm Derek Harris. We're taking a little bit of a detour with this week's episode, at least in the sense of subject matter. Yes, we're still talking about AI, but rather than digging into the cutting edge research around foundation models or sitting down with the founder who's building a product atop LLMs. We're talking about the application of AI to biotechnology and health care overall. And to educate us on this topic, I'm joined by a16z bio general partner Vijay Pande. We start by discussing the folding at home project that Vijay launched back in the year 2000, and follow that through line of distributed computing and massive compute power up through today. Vijay explains how biotech, healthcare, and computer science have evolved together over the past couple of decades, and why we have a golden opportunity right now to accomplish some amazing things and really use AI to improve the quality of life for everyone on the planet. I thought it was a very insightful discussion, and I think you will too.
As a reminder, please note that the content here is for informational purposes only, should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any a16z fund. For more details, please see a16z.com/disclosures. So before we get into where everything is headed, let's rewind a couple of decades, 24 years to be exact. When you launched the Folding at Home project, can you talk listeners through that project and what you're trying to accomplish with it?

**Vijay Pande** (2:11)
Yeah, so let me set the stage. 24 years ago is a long time ago, and computers are, especially like GPUs, are doubling roughly every year in speed. And so like 10 years a decade is like a thousand times of computer power. And so two and a half decades is like a million, a couple million fold of computer power. It's just like mind-blowing, the difference between then and now. And what our goals were then was that I had a central thesis is that we could do a lot at the intersection of computation and biology. We were just basically outgunned computationally, and that there's only so much that amazing algorithms can do, and that compute was a key missing element. And the question was how to get that compute at first. You know, in principle, if I happen to have $10 billion in my pocket, I could whip that out and buy a whole bunch of computers. That was not an option for me. And so what we did is that we built software such that people around the world could donate their computer time.
And we've made several milestones, several supercomputer milestones, such as the first to a pedal flop, which actually is a Guinness World Record right there behind me for that. Basically, we were able to build that supercomputer of sorts in a very unconventional way that was able to, in a sense, propel us forward in time. That it let us do things 10 years ago, 20 years ago, that now are routine, but then was impossible.

**Derrick Harris** (3:37)
Yeah, so you mentioned you didn't have access to all the money to buy, right? And at that point, it would have been like buying a super computer.

**Vijay Pande** (3:42)
Oh, it would be beyond that. It would be something that had never existed before. So it's nothing you would have to create a new computer, essentially. Yeah.

**Derrick Harris** (3:50)
Yes. So so absent that computer explained how you went about accumulating those resources.

**Vijay Pande** (3:55)
Yeah. So so this is also to set the stage. I think one of the big influences on me at the time was Napster. So I started building Folding Home in 1999 and thinking about it. And Napster was like raging then. And I think the Internet was pretty new in 99