self-driving-labs

Self-driving labs used to run the experiment a human designed. The 2026 shift is subtler and bigger: the agent now decides which experiment to run next — and the bottleneck moves to physical validation. That reframes the frontier as compressing wet-lab rounds and building a safe, standard way for reasoning agents to drive real instruments — exactly the layer our lab builds.

A step back from the papers: the ecosystem forming around AI-for-biology. The money is real but pooling at the top and thinning out below, regulators are writing the rulebook this year, and ‘self-driving labs’ are genuine in narrow domains but nowhere near replacing biologists.

Today in AI for life science: Anthropic bets on agent infrastructure over bigger models, the virtual cell meets honest benchmarks, self-driving labs go mainstream, and spatial omics gets its foundation models — plus a new chapter from the lab.