Eli Lilly recently hit a $1 trillion valuation without AI
Lilly became a $1T company without using any AI for the discovery of their blockbuster drugs. With AI they could easily 10X that valuation.
Nvidia and Eli Lilly are building a $1B lab for AI drug discovery in SF.
Lilly CEO Dave Ricks talks about their data advantage: “Some scale [tech-bio] players…are just training on public data but they’re only 4,000 ever approved drugs. Lilly alone has 3 million failed drugs.”
No Digg Deeper questions have been answered for this story yet.
@tunguz Do molecules have better efficacy when discovered by AI?
Is the discovery process a big part of the cost? Is discovery a limit to development spending?
@tunguz Correct. Lilly is about to become the NVDA of healthcare. They have the core platform, manufacturing, and distribution.
It’s still early days.
@tunguz if ai enablement does something is to lower the cost of search and discovery. it's ever more so as new models come out with lower propensity to hallucinate.
@Object_Zero_ Yes
@tunguz Golden age for medicine
@tunguz Yeah, I think it's probably going to 5x pretty soon at least.
@tunguz Errr. Tying GIP in a backbone with GLP-1 while keeping the resistance to dipeptidyl peptidase 4 involved a ton of generative drug design. Is that AI? You could argue so.
@tunguz really depends whether they can capitalize that in products, i can foresee investors getting more skeptic when they see "ai" in business plans
@tunguz ML for genomic analysis is pretty dank.
@tunguz Or they may loose everything to tokens
@tunguz Yes. 10X it. The best commentary is always the 10X commentary. But really, you know that 100X is the new 10X don't you?
@tunguz I think traditional pharma and healthcare industry indeed are titans in the room and all startups want to sell them, but I think longevity direction could be the next new big market ..share my thoughts about it and our experience https://ramp.com/data/ai-jobs-impact
Ever since the discovery of DNA and protein encoding, men have spent their entire lives just figuring out how ONE protein folded after DNA production.
You used to be able to get a Ph.D. just by demonstrating how ONE protein worked. 60 years of work, and we had maybe a couple of hundred proteins worked out.
Between 2020 and 2021, AI discovered how to fold every protein known to man in just 18 months. That entire field is now completely mapped. No more Ph.Ds.
AI will greatly accelerate how quickly we can identify better drug regimens.