Good news!
"Wednesday marked the launch of two notable AI models that promise to accelerate scientific discovery.
The first, Google’s “Co-scientist,” is intended to assist scientists by mirroring “the reasoning process underpinning the scientific method” to generate likely and novel hypotheses. In one test, a team from Imperial College London gave the model a prompt about an antibiotic resistance problem they’d been working on for a decade. Within 48 hours, the AI had not only replicated their leading hypothesis but also provided four plausible alternatives, one of which was completely original.
The second, called Evo 2, was designed by the Arc Institute to read, interpret, and even design entire genomes. This is a big step up from previous biology-focused AIs, which were limited to individual proteins or smaller genetic segments. In tests, the model accurately predicted whether certain genetic mutations would cause cancer—a feat that can take human researchers months."
"[team from Imperial College London] gave "co-scientist" ... a short prompt asking it about the core problem he had been investigating and it reached the same conclusion in 48 hours. ...
given [this] research was not published so could not have been found by the AI system in the public domain. ..."
I just blogged here about Evo 2.
Accelerating scientific breakthroughs with an AI co-scientist (original news release)
No comments:
Post a Comment