Machine learning against biological aging
summary
Claims concrete wins and partial prototypes: blood aging clock R^2 ~77% for chronological prediction, C. elegans lifespan extended ~10x and mice ~100% in lab studies, and porting whale DNA repair genes raises cell radiation resistance ~50%. Uses explainable ML (LightGBM + SHAP) to map features, and builds a retrieval‑augmented LLM stack with vector DBs, agents and embeddings — an open‑source prototype called Longevity Genie that yields better, referenced gene explanations than vanilla ChatGPT but has not yet validated suggested genetic edits in vivo.
AI-written summary · about
chapters (12)
- 0:00Biological aging basics
- 6:56Need for faster aging measures
- 8:00Aging damage and escape velocity
- 9:48Biological age biomarkers
- 11:38Physiological aging biomarkers
- 15:15Explainable AI for aging clocks
- 19:34Longevity secrets from species
- 23:11Genomic insights in longevity
- 26:49Longevity Genie with AI
- 31:52LLMs for biological data
- 37:10AI safety in biology
- 40:44Open-source aging research
related talks
21mYounger Tomorrow! The Age of Longevity
20mHow not to die
38mBig data in personalised medicine
34mThe True Revolution Lies Ahead of Us
30mMachine Learning: A New Approach to Drug Discovery
39mThe Next Frontier of Longevity: Harnessing science & technology for health & wellness
20mAdvancing Epigenetic Predictors with Scalable Machine Learning
28mThe Algorithm of Grief: From Loss to Legacy