Analysis
Astromech, an AI company built to model evolutionary biological change, raised new funding, SiliconANGLE reported Thursday. Biotech investor Bob Nelsen led the round, with Peak6, NeoGenesis Capital, Builders VC and CAZ Investments participating.
- New raise -- $20M
- Valuation -- $3.8B
- Total raised to date -- $60M since founding
Colossal's second act
Astromech was co-founded by Ben Lamm and George Church, the same pair behind Colossal Biosciences, the de-extinction company known for its work toward reviving traits of the woolly mammoth and dire wolf. Astromech was gestated inside Colossal before spinning out as its own entity, and it applies a related but distinct thesis: rather than editing genomes toward a specific historical target, Astromech is building a predictive model of biology itself, using genomic, evolutionary and functional data spanning roughly 3.8 billion years of evolutionary history -- the figure the company's valuation nods to directly -- as a training signal.
The company describes its goal as forecasting how living systems may change and where biological systems are vulnerable, drawing an explicit comparison to weather forecasting: rather than predicting a single organism's fate, Astromech's models are meant to recognize patterns of biological resilience and vulnerability across species and time, which the company argues could eventually inform drug discovery, pandemic preparedness or agricultural resilience, though none of those applications are yet productized.
Competing in a crowded AI-biology field
Astromech enters a category already populated by well-funded players: Google DeepMind's AlphaFold franchise established protein-structure prediction as tractable years ago, and Anthropic's own Claude models recently demonstrated functional protein-binder design in an external lab test with a hit rate roughly double the industry norm. Isomorphic Labs, Xaira Therapeutics and Chai Discovery are all running their own AI-driven biological design pipelines with pharma partnerships attached. Astromech's differentiation, on the evidence available, is scope: rather than a narrow task like binder design or structure prediction, it's positioning itself as a broader "biological operating system" for forecasting evolutionary change generally -- a more ambitious and less immediately falsifiable claim than a specific lab-validated result.
The numbers in context
A $3.8 billion valuation on $60 million total raised and no disclosed revenue is a striking multiple even by the standards of 2026's richly priced AI-biotech sector, and it leans heavily on Lamm and Church's prior track record -- Colossal itself has reached a multibillion-dollar private valuation on largely pre-commercial science -- rather than any demonstrated Astromech product. The new capital is earmarked for expanding the research team, broadening the number of species represented in its comparative genomic datasets, and scaling training infrastructure, all of which are inputs to a better model rather than evidence one already exists at the scale the valuation implies.
Investors underwriting this round are betting on the founders' capacity to execute a genuinely novel scientific thesis before any competing effort -- from DeepMind, Anthropic-adjacent life-sciences work, or a pharma-backed rival -- reaches a comparable predictive capability first.
