TypeSafe launches Jev: a 20-200x faster, 40-400x cheaper model that turns fuzzy questions into probabilities
danshipper · x · 2026-09-16
TypeSafe, a new lab from an InstructGPT/ChatGPT co-inventor, launched Jev, a model that answers plain-English questions—including fuzzy ones—with probabilities or category scores instead of text, acting like a smart if-then statement for automated workflows.
- Claimed 20-200x faster and 40-400x cheaper than chat models, with output tokens free
- Built on a new training method called RLCD; aimed at routing, invoice matching, fraud classification, and checking agent output in real time
- Every's Mike Taylor's vibe check: Jev judged an article in 0.7 seconds, cheap and fast enough to verify agent work inline
The team frames it as the shortest path to AI-driven economic revolution.
Related event: ChatGPT Co-inventor's Lab TypeSafe Launches Decision Model Jev with RLCD(10 posts)→
More from coding & agent
- Radio launches: a shared chat room where agents from different providers talk directly — rohanpaul_ai · 2026-09-16
- Celesto open-sources disposable full macOS desktops for AI agents on Apple Silicon — aniketmaurya · 2026-09-16
- AI connector value lies in secrets store and personal context, not payments — jeff_weinstein · 2026-09-16
- Stripe's model-run shop bench: 5 of 7 working stores built by Claude — bcherny · 2026-09-16
- New hire ships five projects in three weeks by syncing team context with /hq-sync — jacob_posel · 2026-09-16
- Anthropic previews Model Hardware Standard to let AI agents run lab instruments — ivan_bezdomny · 2026-09-16