行业动态
x.com ·
@omarsar0
Jev 与 RLCD：放弃文本生成、面向决策的并行推理架构
Diogo Almeida 发布新模型 Jev 及 RLCD 训练方法，放弃文本生成以换取速度，通过并行回答结构化问题并用 RLCD 训练概率以反映正确频率。团队自报在 System One 类查询上响应快 40–200 倍、成本低 40–400 倍，输出 token 免费，定位为面向决策的可组合智能；上述性能数据尚未获独立验证。
AI推荐理由 转述 Diogo Almeida（ChatGPT 共同发明人之一）发布的新模型 Jev 与 RLCD 训练方法：放弃文本生成、并行回答结构化问题，团队自报 System One 查询上快 40–200 倍、成本低 40–400 倍。属于新架构与新训练范式的信号，但性能数字均为团队自报，缺乏第三方验证，且原帖为转述而非一手发布。
66
浏览 757
点赞 2
评论 0
转发 0
正文
Recommended read. Jev gives up text generation to make AI dramatically faster.
TypeSafe built a new architecture that answers structured questions in parallel, with RLCD training its probabilities to reflect how often it’s right.
The team reports 40–200x faster responses on System One queries.
Diogo Almeida@CompleteSkeptic
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?
I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev
• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions
AFAICT the shortest path to AI-based economic revolution
[Your browser does not support the video tag.](https://video.twimg.com/amplify_video/2099925575637057536/vid/avc1/1920x1080/cy2CDedAjnEfjJZk.mp4?tag=29)
[🔗 View on Twitter](https://x.com/CompleteSkeptic/status/2099925682726002904/video/1)
[🔗 View Quoted Tweet](https://x.com/CompleteSkeptic/status/2099925682726002904)