Cryosphere Podcast Mark Woodward - A Revolution in Cryoprotectants
Apr 6, 2025
Mark Woodward, founder of Wake Bio and former AI researcher at Google Brain, dives deep into the world of cryoprotectants. He reveals how machine learning is being used to discover chemical combinations that protect biological tissues during freezing. Starting with zebrafish, Mark's ambitious goal is to achieve whole-body human preservation. He discusses the challenges of transitioning from lab to reality, rapid warming methods, and the ongoing debate between biological reanimation and uploading. With a five-year target for preserving adult mice and a ten-year vision for humans, the future of cryonics is looking bright.
AI Snips
Chapters
Transcript
Episode notes
Run ML-Guided High-Throughput Screens
- Prioritize high-throughput experiments guided by ML to scan chemical compositions efficiently.
- Use models to propose promising cocktails before wet-lab testing to save time and resources.
Limited Exploration Explains Few Breakthroughs
- Cryobiology has explored only a narrow chemical space; M22 largely reflects limited historical search effort.
- A small field and slow traditional assays explain why major improvements since M22 are scarce.
Balance Toxicity, Viscosity, Permeability
- Optimize multiple properties: low toxicity, low viscosity for perfusion, membrane permeability, and high glass transition.
- Aim for rapid glass transition at practical temperatures while keeping perfusion feasible.