TypeSafe, founded by a ChatGPT researcher, launches Jev — a decision engine that can't write

Scobleizer · x · 2026-09-18

TypeSafe, a lab started by one of the researchers behind ChatGPT, released its first model Jev: a decision engine that takes options, picks one, and reports its confidence — it cannot write a single word. The bet: when software needs a plain yes/no, forcing a model to write a paragraph and parsing the answer out is the wrong interface; the valuable half was never the writing. Scoble reports teams like Sully.ai are already building with it, 48 hours after launch.

Related event: Self-Proclaimed ChatGPT Co-Inventor Launches Decision-Only Model Jev(4 posts)→

Original post →

More from Models

Models channel →