01:42:18 · John Coogan
[01:42:18] S1: I I I remember what was it? In, like, 02/2008, there was a boom in electronic medical records startups like Athenahealth was one of them. And there was kind of like two waves going on that were kind of adding together and making those businesses work. One is like a technological wave obviously like we had access to cloud databases and cloud infrastructure and better software and so you could actually like put medical records securely in a cloud and so that was obviously like a great business. But then there was also an EHR mandate from the Obamacare legislation I believe. Are there any regulatory tailwinds that are adding on top of, like, the AI technology tailwinds right now in bio or health Yeah. Care
01:43:02 · Vineeta Agarwala
[01:43:02] S7: Absolutely. So there are first of all, I will say, historically, historically, a lot of health care software and technology was adopted because of regulatory incentives and payment incentives. Yep. But for the first time, we are seeing bottoms up adoption of tools and technologies and and an extraordinary NPS of health care software with scribing tools, like those developed by companies like Ambience and Abridge, you know, you know, you know, evidence, access tools like what what OpenEvidence has built. And so for the first time, we're also seeing pull that's just like, hey. This is better. I don't need the government to give me a, you know, a kickback to tell me it's better, but it is better. So that's a separate trend that we're very excited about. But huge regulations are also going to incentivize a better technology infrastructure. So price transparency is a good example. There are clear incentives and penalties for not being able to share transparently your price your pricing cards as a provider, as a payer, and to create that type of open access for patients. And so building that infrastructure layer is gonna be super important. The administration's talked a lot about patient choice. That could be patient choice with respect to how you use your benefit dollar, patient choice with respect to what plan coverage you select. And ICHRA is a good example of a potentially big change in the employer sponsored health care space. So these are all financial rails and incentives. CMMI continues to advance value based initiatives that basically prioritize preventive care and say, hey. At the end of the year, if you save the government a dollar, you save CMS money, dear smart provider who's using technology, you'll make money. Right? So that's a clear financial incentive. It's good for patients. Generally, being out of the health care system, preventing something before it costs a lot of money is good for patients. But the financial incentives and policy incentives do need to, you know, do provide a huge accelerator in our space, and that will continue to be the case.
01:45:08 · Jordi Hays, Vineeta Agarwala
[01:45:08] S0: Can you give us the update on the drug pricing EO from, I don't know, was it six weeks ago, eight weeks ago? It feels like a year ago at this point. There's been a lot that happened. We had a bunch of people on the day of, and a lot of people were kind of saying that it wasn't going to be as impactful as it kind of felt in the moment, but what what's the update there? [01:45:30] S7: I don't think there's been a policy update. Like, you know, nothing has been enacted yet. That's still an EO. Yep. The the concept of it, of course, was to say was to use international trade as a backdrop for holistic negotiations inclusive of drug pricing to say, you know, as part of an overall tariff negotiation, why is it the case that The US pays two or three times or more for the same exact drug that another another country pays? And so that it is a logical question to ask, and, we know that CMS is continuing to ask that question. But the actual policy is gonna depend on quite a bit of, you know there are a number of moving parts and a holistic policy with respect to every other every other country. So it's important for us to retain huge incentives, and it's not gonna be as simple as cut the price, get great drugs. That's not how the market works. Yeah. But should the price be more equitably shared across countries, it's an absolutely reasonable perspective to have.
01:46:36 · John Coogan, Vineeta Agarwala, Jordi Hays
[01:46:36] S1: How popular is open source in the bio world? I feel like the Cursor for bio thing is very interesting. But when I think about the electronic lab notebooks that folks are using, like, if it's not open source, you can't fork it. And it feels like part of Cursor's ability to grow really fast was that they they sit on top of your code base and you can change IDEs very quickly. But if you're you if a researcher is using a closed source lab notebook that has a database built in, that company is gonna be like, no, no, no. Like, I do not want you just stealing all my customers or like migrating off. But are is open source popular? Are there opportunities? Like, what's the VS Code equivalent, I guess, would be the question? [01:47:23] S7: Yeah. I think there's a lot you know, in bio research, there's a lot more informatics and and less organized software. I think that will change. But historically historically, there is a lot of bespoke informatics, which is why some of the agentic adoption, like, for exploratory informatics research is happening on its own, and people are adopting open source agents for resources like that. You know, the structure the protein structure models like AlphaFold [01:47:53] S0: Sure.
01:47:53 · Jordi Hays, Vineeta Agarwala
[01:47:53] S0: Sure. [01:47:53] S7: Were followed by a number of models that were open sourced. And within a period of weeks, thousands of scientists, thousands of companies adopted, additional structure prediction models. Mhmm. Boltz two was a popular one. The Chai model. A number of startups put out open source models, and we saw really fast adoption, which is kinda cool. Right? It's like it's a scientist sitting in a company who says, I wanna try that, you know, and see if it would help my workflow. And, actually, I wanna benchmark it, and I wanna try three at the same time. So we didn't have that many great open source resources like that before we had datasets, but not necessarily models and APIs to those models that made them easy to use. So I think that that in that whole ecosystem is in its, like, square one infancy of growing and exploding, but you're gonna need a library of models. You want to be able to try all the models. You want your agents to be able to use the models without even a human interface. And so suddenly, the same thing that's happening in enterprise software in every vertical is gonna happen here too. You know? Agents on behalf of scientists will be able to select which models to use, try five models at once, and then incorporate the results and predictions of those models into a workflow, you know, that is deeply integrated with, you know, with other agentic automation. So it's none of that's built, I would say, for for the vast majority of the industry. And so it's a huge, huge opportunity for startups to build.
01:49:25 · Jordi Hays, Vineeta Agarwala, John Coogan, Casey Neistat
[01:49:25] S0: Yeah. Last quick question. Maybe if you could answer in in sixty seconds, that's the challenge. What what is the, what's the vibe on the Stanford campus right now? Are people dropping out of, you know, PhD programs to start companies in in bio? I'm sure we're seeing, people dropping out of PhDs to join meta and other AI research orgs broadly, but I'm curious if that's happening in in bio. [01:49:53] S7: There are a number of fast growing AI biotech startups, right, that are bringing AI and computational tools to model training. So we are seeing that. If you're asking, know, is it a response to federal funding changes? I think the you know, that's always been the case that a large fraction of trainees from university campuses end up in biotech and pharma companies, and that's great. It's a great outcome. I don't think you know, I do think there's probably, in the near term, going to be continued exploration of industry alternatives to federally funded research. And I really hope that there that the the, you know, federal grant funding is clarified and and and and that there's more stability on university campuses for federally funded research because some of that infrastructure is absolutely has been behind [01:50:45] S0: Shaking. [01:50:46] S7: A huge number of these of these innovations. [01:50:49] S0: Makes sense. Thank you so much for joining. [01:50:51] S1: Yeah. You [01:50:52] S0: so much. Great to chat. [01:50:53] S1: Talk to you soon. Cheers. [01:50:54] S6: Thank you.