用 Jev 做决策层：论文分类、邮件筛选、航班搜索实测
把「创造交给 LLM、选择打分交给 Jev、执行交给代码」拆开，公开 demo 显示千篇论文分类约 0.08 美元、500 封邮件约 3.5 美分。
this JEV is straight up f*cking diamond every agent you ship wastes frontier tokens on questions that were never writing jobs is this spam? which worker runs next? is the draft ready to show a human? those are decisions. Jev is built for exactly that half of the loop you pass in state plus a typed question. it gives back a choice, a score, or yes/no with confidence. no chat. no second opinion the public demos, not our lab: > ~1,000 papers classified for about $0.08, median around 256ms > ~500 emails in a single inbox pass for about 3.5¢ > a flight search loop in roughly 7 seconds for about $0.0039 > a huge Claude session compacted from ~1M tokens down to ~86k in about a second the whole trick is the split if it creates, the LLM writes if it picks, scores, or gates, Jev decides if it does the thing, code runs it pricing people quote: around $0.042 per million tokens in. output is free. 10,000 decisions comes to about $0.42 a local router makes it obvious. you give it the goal and what's already done. Jev picks research, write, or review. anything under 0.85 confidence goes to a human. the job lands as JSON in a folder. the writer never wakes up for a yes/no and the loop won't publish on its own drafts go to review. you hit post. Jev never sends the tweet one model writes. a cheaper one decides. nothing irreversible ships until a human says yes
这条是引用着另一条写的：x.com/asteri_eth/status/2104901781725598103