Bryan Johnson's Sleep AI Identifies You With 92.5% Accuracy From One Night
iScienceLuvr · x · 2026-09-05
Bryan Johnson teamed up with sleep researcher @mfranceschetti to launch an AI model trained on raw sleep biosignals, now live in the Eight Sleep app.
Capabilities: it can identify you from thousands of users with 92.5% accuracy from a single night's signal (a "sleep fingerprint"); predicts biological age within 3.3 years — claimed to be the most accurate contactless bioage estimate ever reported; detects diabetes better than Apple's model; and flags heart failure at AUC 0.822 plus speed of aging.
The training set is billed as the largest raw biosignal dataset ever used for an AI model from a single device (hundreds of thousands of hours). Commenters note self-supervised learning in medical AI remains badly underrated.
More from Models
- GPT progression: GPT-3 text, GPT-4 images, GPT-5 video, GPT-6 3D? — PaulYacoubian · 2026-09-05
- Open models should chase depth, not Claude-style coding, argues researcher — teortaxesTex · 2026-09-05
- Reddit user claims Astra's capability jump exceeds GPT-3.5→4, yet the internet is silent — Glittering-Neck-2505 · 2026-09-05
- GPT-6 Astra claimed as new SOTA in agentic CAD — mckbrando · 2026-09-05
- GPT-6 hyped as AGI can't even render a snake with its head attached — SnooCheesecakes1893 · 2026-09-05
- GPT-6 Astra usage limits hit every 5 minutes, user can't even reset anymore — memeka · 2026-09-05