TypeSafe AI launches Jev, a 'System One' model for bounded decisions in agent runs
hardimanjames · x · 2026-09-25
TypeSafe AI introduced Jev, its first "System One" model targeting a task we currently force generative LLMs to handle: bounded decisions inside software. The accompanying article argues much of an agent run isn't reasoning but a stream of small decisions — which model handles the next step, whether a tool call needs approval, whether the agent is still making progress, whether the task is done. Today these go to another generative LLM, which still generates tokens sequentially and constrains them into a schema before the app converts output back into a decision. Jev aims to remove text generation from that loop.
More from coding & agent
- TypeLLM adds type-safe, schema-guaranteed generation to LLMs without touching weights — kalyan_kpl · 2026-09-25
- Five mistakes people make with agentic factories: micromanaging, no quality bar, and 'puppet shows' — hugobowne · 2026-09-25
- Open Claude Code skills catalog ships 31 testing skills, from metamorphic testing to E2E — doodlestein · 2026-09-25
- Researcher proposes agent-suggests-human-executes loop for real-world science experiments — suragnair · 2026-09-25
- CrawlRaven hits 700 users in 60 days, targets 2,000 more with MCP-powered SEO audits — ayushtweetshere · 2026-09-25
- Human-in-the-loop or machine-executed: verifiable tasks turn agent traces into RL rewards — suragnair · 2026-09-25