Mollick argues AI labs may enter finance and biotech on internal-model gap
Article timeline
The articles in this story. Times are ET.
Ethan MollickXMollick suspects a growing gap between lab-internal and public AI modelsAIEthan Mollick says there is no good measure of the gap between the best internal models at AI labs and the best publicly released model. He suspects the gap has grown over the past year, especially after the temporary Fable ban.
Egan PeltanXFrontier AI labs are moving into biotech to speed up drug R&DAIOpenAI acquired Orchestra, an operational platform for drug programs, and Anthropic released vignettes on Claude helping teams accelerate drug development along with its acquisition of Mithril Medicines. The author argues the labs' motivations are not primarily financial but aim to prove new human-AI collaboration methods for physical-world productivity. The post notes that drugs deliver lasting benefits, while drug development takes years, as with Eli Lilly's retatrutide, which entered Phase 1 in 2019.
Andrew CurranXOpenAI, Anth, and Google are all leaning heavily into bio. I know people are skeptical about results in this area, and I understand why, ...AIbut I honestly believe that if you are reading this, there is a non-trivial chance that (if you want to) you will live for a very long time.
- Ethan MollickXAI labs may enter finance and biotech using internal model advantages
AIEthan Mollick argues that as labs' internal models pull ahead of released ones due to safety concerns and research advantages, OpenAI and Anthropic will be increasingly tempted to exploit that gap by entering new markets, especially finance and biotech.