“The real bottleneck in drug development isn't sort of target identification, drug design. The bottleneck is clinical development. And I just don't know outside of on the margins of efficiency how you're going to get around testing a year and a half or an outcome-based study in a cardiovascular disease in 10,000 patients that'll take five years to run and cost $300 million, right?”
sputnikintelligence.com
The hosts were debating how far AI will change biotech after Anthropic's S-1 flagged life sciences as a strategic push. Skorney pushes back on the idea that AI will transform drug development, arguing the limit is human trials, not discovery.
…better? And I think, again, in some areas, it will do. I know, again, companies that are using it specifically for that purpose based on population data, based on biobank data. So I'm very excited by it. But, you know, the fruits will take time because you still have to put them through humans.
…of the day, especially new targets, you won't know the rubber hits the road with a true clinical study in human subjects, right? And a data readout. It'll be interesting to see how that evolves. And if we increase the success rates, you know, Greg, Brian, anything else to comment on the AI theme?
Yeah, I mean, I was having dinner the other night. We were talking about this. And, you know, a PM I was discussing it with was like, oh, this is definitely the most transformational technology that we've seen in our lifetime. I kind of pushed back and I was like, well, I don't know. I mean, it's super transformational. We use it all the time. But I would argue like you can take a lot from the internet itself was super transformational, right? I mean, we've sort of democratized information flow, but there's a lot of problems that have come with that too, right? There's a ton of information on the internet. There's a ton of misinformation on the internet, right? And AI functions as sort of a way of taking in a lot of information and analyzing it very, very quickly. But I would assume like myself, everyone on here who has been using AI finds, it makes a lot of mistakes too, right? And it's probably because, you know, like a lot of people in the tech world use the term Geiger, which is garbage in, garbage out, right? So, you know, getting information in that's right and analyzing it and getting the right outputs is very important. But Chris, you bring up a very good…
Yep. Yep. Greg, any final comments on
that? Yeah, I would just argue
that I do think that AI will be transformational for everyone, but particularly in biotech. But to Brian's point, clinical development is still just really the biggest challenge. I'm not sure how AI will necessarily speed up mid to late stage clinical trials. But at the same time, I'm hopeful where…