Edison Scientific announces partnership with Incyte to deploy Kosmos AI agent across full drug development pipeline from molecular design through FDA approval

SRSam Rodriques@SGRodriques

We live in a golden age of biology. So why are people still dying from disease? Because discovery and development move slower than they should. Today, we’re partnering with Incyte to change that. Kosmos is now the first agent that can compress months of drug development into weeks, from the earliest stages of scientific discovery through to FDA approval. @Incyte will be the first company to deploy it across their pipeline. Work that used to take a team of scientists months now happens in weeks. Patients can't wait, and neither can we.

CECarlos E. Perez@IntuitMachine

I just fell down a rabbit hole reading a new AI paper and I think my brain is broken. An AI system called Kosmos just did what collaborators estimate is 6 months of PhD-level scientific research. It did it in 12 hours. 1/12 I know, I know. We hear "AI scientist" and think it's just a chatbot that's good at summarizing Wikipedia. I was skeptical too. Most of these systems are toys. They can do a cool analysis, but they lose focus after a few steps. They can't run a real, long-term investigation. 2/12 The real problem wasn't raw intelligence. It was coherence. Imagine trying to write a book with 100 different people who can't see what anyone else is writing. You get a mess of disconnected paragraphs. That's what previous AI agents were like. Brilliant, but hopelessly siloed. 3/12 So the team behind Kosmos didn't just try to build a "smarter" brain. They built a shared consciousness. They call it a "structured world model," which sounds complex, but the idea is genius. 4/12 Think of it like a giant, live-updating whiteboard. Kosmos unleashes hundreds of little AI agents in parallel. One reads scientific papers. Another analyzes data. When an agent finds something, it puts it on the whiteboard. Crucially, every other agent can see the whole board. 5/12 This simple change in architecture lets it scale to an insane degree. In an average run, Kosmos: Writes ~42,000 lines of code Reads ~1,500 full scientific papers (I had to read that twice. It's not a typo.) It lets the system pursue a single goal for hours without getting lost. 6/12 "Okay," you're probably thinking, "but how do you know it's not just confidently making stuff up?" This is the part that actually blew my mind. Every single claim in the final report Kosmos generates is hyperlinked. You can click any sentence and it shows you either: A) The exact Jupyter notebook with the code it ran. B) The primary scientific paper it's citing. It shows all its work. No black boxes. 7/12 And the results are just... wild. In one test, they gave Kosmos a dataset on brain aging that human scientists had already analyzed for years. Kosmos found a completely novel mechanism for why certain neurons are vulnerable—a discovery the original researchers had missed in their own data. 8/12 It wasn't a one-off fluke, either. The paper details 7 major findings. It independently reproduced discoveries from unpublished manuscripts. It developed a brand new analytical method for studying Alzheimer's. It found a potential new drug target for heart disease. This thing is a cross-domain discovery engine. 9/12 This forces a huge mental shift. We've been thinking of AI as a tool to answer our questions or summarize existing information. This is different. This is a tool for generating new, verifiable knowledge from scratch. 10/12 This isn't just about making science faster. It's about changing the kinds of questions we can even dare to ask. We're about to enter an era where a single researcher, armed with a system like this, can explore ideas that would've previously taken entire institutions years to investigate. 11/12 The pace of discovery is about to get very, very weird. If you read one paper this week, make it this one. 12/12