Interview: Medical AI is only as good as human consensus labels
CurieuxExplorer · x · 2026-11-22
Ljubomir Buturovic, ML lead at Inflammatix, discusses building diagnostic models without a gold standard: a perfect score against expert labels doesn't prove perfect diagnosis, since the reference itself is fallible human consensus. He traces his path from an ML PhD in Belgrade through Incyte and Pathwork to Inflammatix, and how an engineering mindset of shipping to specs still shapes his algorithmic diagnostic work.
More from AGI Musings
- Musk says Grok's logo represents the Singularity: unpredictable but very interesting — XFreeze · 2026-11-22
- Why a P=NP proof would collapse most of modern cryptography — doctorBobG · 2026-11-22
- Disney's 1940 Fantasia already depicted the AI alignment problem, decades early — JeremyNguyenPhD · 2026-11-22
- HN debate: Is the AI boom becoming a disaster for the climate? — Brajeshwar · 2026-11-22
- Terence Tao's deeply physical problem-solving sparks debate: does intelligence require embodiment? — CatAstro_Piyush · 2026-11-22
- AI safety researcher urges US-China coordination on a common frontier AI red line — S_OhEigeartaigh · 2026-11-22