What's Your Problem? Can AI Help Solve Alzheimer’s?
9 snips
Feb 5, 2026 Patrick Hsu, co-founder of the Arc Institute and UC Berkeley bioengineering professor, merges AI with experimental biology to speed discovery. He discusses making labs efficient with model-guided experiment design. He explains DNA autoregressive models, AI-designed phages, virtual cells predicting perturbations, and implications for Alzheimer’s and therapeutics.
AI Snips
Chapters
Books
Transcript
Episode notes
AI Can Replace Much Of Guess-And-Check
- Biological research is still manual and slow, relying on months-long guess-and-check experiments.
- Patrick Hsu argues AI models can rank promising experiments and vastly speed iterative lab discovery.
Adopt Model-First Lab Workflows
- Use AI as a design co-pilot to change how scientists read papers and design experiments.
- Patrick Hsu urges adopting model-first workflows to rank experiments before running slow wet-lab tests.
Full-Stack Lab+AI Speed Breakthroughs
- ARC Institute combines experimental and computational biology under one roof to run tight model–lab feedback loops.
- That integration aims to produce productizable discoveries, not just academic papers.
