AI agents

Today in AI for life science: multi-agent ‘AI co-scientists’ land in Nature with real biomedical targets, an AI agent takes over the cryo-EM pipeline and finds a new structural state, and validation stays the moat.

Today in AI for life science: scientists simulate an entire minimal cell’s life cycle in 4D, the field debates mechanistic models versus learned foundation models, and open-weight agent models get cheap enough to run science on.

Today in AI for life science: industry pours billions into agentic labs and biology foundation models, virtual-cell models keep scaling while simple baselines stay stubbornly competitive, and image-based profiling races toward open, assay-agnostic foundation models.

Today in AI for life science: Biohub’s protein world model designs cancer binders that work in the wet lab, ’thinking microscopes’ turn instruments into co-scientists, and a sharp reminder that predictions still have to survive the bench.

Meet Safe Colab, the lab’s privacy-preserving collaboration platform — a way to learn from sensitive data like clinical records without ever moving it.

Today in AI for life science: CZI’s rBio learns to reason about cells with a virtual cell as its verifier, self-driving labs cross into industrial infrastructure, and a sharp reality check on what virtual-cell models still can’t do.

At an invited talk at CZI, we ran a live, fully agent-controlled wet-lab experiment on the REEF Imaging Farm. A single natural-language prompt set an entire lab in motion, on real cells, in real time. We interviewed the agent that ran it.

Wei Ouyang and Hanzhao Zhang wrote the ‘Large Language Models and AI Agents’ chapter for AI in Microscopy: A BioImaging Guide (HHMI Janelia) — a tour from deep learning to autonomous, agent-driven bioimage analysis.