Future of Life Institute Podcast AI vs Cancer - How AI Can, and Can't, Cure Cancer (by Emilia Javorsky)
53 snips
Mar 16, 2026 A clear-eyed look at where AI truly speeds cancer research and where hype falls short. Topics include targeted AI wins like AlphaFold, the data and lab bottlenecks that block clinical progress, and why biological complexity resists software-style thinking. The conversation covers systemic incentives, regulatory mismatches, and a practical roadmap for building data, funding, and policy infrastructure to make AI helpful rather than magical.
AI Snips
Chapters
Transcript
Episode notes
Reproducibility Crisis Threatens Medical AI
- Much published biomedical research is non‑reproducible, poisoning AI training data.
- 70% of researchers report replication failures and 68% of papers lack raw data, risking garbage‑in garbage‑out models.
Scale Human‑Derived Preclinical Models
- Prioritize funding for standardized, high-throughput human-derived experimental platforms.
- Scale organoids, patient-derived xenografts, phase-zero microdosing, and robotic automation to improve human predictive value.
Optimization Culture Harms Biological Systems
- Narrow metric optimization prevalent in tech can harm biology where balance matters.
- Driving single biological metrics to extremes often causes harm; AI optimizing wrong incentives risks worsening care.