LatchBio on AI in Biology: Frontier Models Untrustworthy Without Verifiable Benchmarks
AI Engineer · youtube · 2026-08-01
In a talk on verifiable environments for AI in biology, Kenny Workman of LatchBio highlights that a single spatial biology run yields 2-6 terabytes of data, providing massive raw material for AI training.
His team adapted coding models into biology tools and built sequencing-based benchmarks, finding that current frontier models cannot yet be trusted with these complex tasks. Workman emphasizes designing tasks that reward reasoning over memorization. He also shares trajectory data from real scientists and candidly discusses biosecurity challenges, including model refusals and red-team tasks.
Related event: LatchBio Warns Frontier AI Models Remain Unreliable in Biology(2 posts)→
More from Research
- Transforming quadruped robot runs on real hardware: per-pose RL policies in Isaac switched on morph — _Stocko_ · 2026-10-16
- KinoGen synthesizes robot loco-manipulation reference motions directly via kinematically grounded video generation — ChongZzZhang · 2026-10-16
- Science Robotics paper introduces F³, predictive event-based vision beyond optical flow — CSProfKGD · 2026-10-16
- VideoGen-Agent: An 8B Multimodal Agent Turns Video Generation Into a Tool-Use Loop — jiqizhixin · 2026-10-16
- Comprehension Audits: New Paper Proposes a Gate for AI-Written Code at Frontier Labs — ronbodkin · 2026-10-16
- Not Boring #215: Cammarata's Clean Networks Claim, Tesla's Flying Roadster, Tau Cut 73% — Not Boring (Packy McCormick) · 2026-10-16