ChatGPT co-inventor launches Jev, claims 20x faster and 40x cheaper
multiply_matrix · x · 2026-09-18
An account claiming to be a ChatGPT co-inventor unveiled the result of two years in stealth: a new training method (RLCD) and a new frontier model, Jev, claiming 20-200x faster, 40-400x cheaper (with free output tokens), optimized for composable decision intelligence.
Researcher Andrew Ratner's reshares highlight another angle: the comeback of discriminative/predictive models — he has long argued that in settings like closed-schema classification, classical ML formulations beat generative chat on accuracy, speed, and typing.
More from Models
- GPT-6 Astra One-Shots $10k Gambonanza Challenge in Obscure Game, Fully Autonomous — Vjeux · 2026-09-19
- JEV classifier sparks buzz: the moat is the data, not the model — MoonL88537 · 2026-09-19
- A Live Open LLM Pretrain and RL Training Run You Can Watch in Real Time — mariofilhoml · 2026-09-19
- tokenbender: the long-term path is weak experiential learning, not pure — tokenbender · 2026-09-19
- ChatGPT Pro 20x accounts so scarce that a black market via GCS has emerged — almmaasoglu · 2026-09-19
- MiniMax teases upcoming H3 model in official repo — const_reborn · 2026-09-19