AI × Science

We make biology controllable

by training useful general scientific intelligence. Our North star is to solve aging by making the process of therapeutics discovery and development a commodity

Our approach Get in touch →
Three panels: the Bitter Lesson — a monstrous datacenter chewing through raw data; the scientific loop of data, hypotheses, experiments, Occam's razor, theory and generalization; and science as compression, where scattered data is distilled into a compact theory that generalizes.
Team

From the team behind

50× improvement of cellular reprogramming by AI protein engineering

within Retro × OpenAI collaboration

25% of deals between aging biotech and pharma

among 2020-2025 partnerships according to age1

$1B+ in negotiated transactions

including Gero × Chugai deal

Founders

Approach

Science isGenerative Selection.

The scientific method is a loop — generate hypotheses, test them against reality, and select what survives by Occam's razor. We rebuild that loop with AI and a wet lab: compressing data into knowledge, not just scaling compute.

01

Generate

AI poses productive problems and proposes testable protein designs — installing priors to collapse the search space

02

Test

We run the highest-information-gain experiments in our own wet lab, closing the loop in days — not years

03

Select

Reality selects the designs that are optimal for a desired function. Each result sharpens the next generation — signal over scale.

First application: speeding up the design of biologics to optimize a desired function with the use of AI. Our north star is aging.

Let's build it