TypeSafe ships Jev: typed calibrated decisions, no RLHF and no text generation

ccerrato147 · x · 2026-09-20

Thread conclusion: the fix needs a different objective — not human preference (RLHF), not verified correctness (RLVR), but calibrated decisions: confidence tracks accuracy, so software acts when sure and escalates when not. Diogo's TypeSafe shipped it as Jev on Sep 15: typed decisions, no text generation.

Related event: ChatGPT co-author argues RLHF optimizes the wrong goal; TypeSafe ships Jev for calibrated decisions(9 posts)→

Original post →

More from coding & agent

coding & agent channel →