Jev 正式发布:面向决策的 System One 模型
TypeSafe 创始人 Diogo Almeida 介绍 Jev、RLCD 训练方法,以及面向软件决策场景的模型定位。原帖附有 2 分 56 秒发布视频。

曾在OpenAI构建指令遵循模型的人现在运营TypeSafe实验室,推出Jev模型。该模型不生成任何文字,只做决策。他在发布后花了两个多小时解释原因。
The man who helped build instruction following models at OpenAI says the next wave of AI won't look like chat. Now he runs TypeSafe, the lab behind Jev. A model that never writes a word. It only decides. He spent over 2 hours explaining why, days after launch. Here are the