X

读了一篇叫HeavySkill的论文，非常有意思。 让多个 AI先并行“独立思考”，生成多条独立推理。 再用另一轮推理来综合所有思路，得出最终答案。 按论文测试结果，回答质量会提升非常多。 正在按这个思路写一个Skill，Claude Code推理，Codex当主持人，看看效果。 blog.qiaomu.ai/heavyskill-he…

For Codex, we’ve been thinking about keeping a stable release cadence and have a larger release each week on Thursday. That does make the start of the week and bit less exciting. Thoughts?

我给每个下属都配了一个专属 Agent，跑在飞书上。现在是他们的 Agent 在跟我的 Agent 对话，我在旁边看着。 带团队这些年，我最大的感受不是累，是碎。 各种项目要跟，各种进展要盯，各种需求要确认。AI 让每个程序员的产出翻了好几倍，活多了，事也多了，人还是那几个。 有一天我看着 Claude Code 自己把一个功能从需求写到上线，然后转头看了眼飞书群，纪要要人整理，进展要人跟，审批堆着等人看。开发层已经 AI 化了，协作层还是原始人。 正好发现飞书 CLI 过去一个月悄悄更新了 100 多条能力，很多上个月还不存在的东西，现在已经可以用了。我就开始动手改。

🚨 There's a major attack going on via npm right now. Do not install any packages right now. Talk to your agent ASAP and check if you're vulnerable or have been compromised. This is urgent ‼️

读了一篇叫HeavySkill的论文，非常有意思。 让多个 AI先并行“独立思考”，生成多条独立推理。 再用另一轮推理来综合所有思路，得出最终答案。 按论文测试结果，回答质量会提升非常多。 正在按这个思路写一个Skill，Claude Code推理，Codex当主持人，看看效果。 blog.qiaomu.ai/heavyskill-he…

大家想读什么论文或书，但不想消耗自己Token。 欢迎留言，我来帮大家跑，解读成图文并茂的文章，一起学习。 示例为ViT经典论文解读。 blog.qiaomu.ai/vit-vision-tr…

nitter.net/UnitreeRobotics/status… At this point on, I'm too afraid to ask what Unitree's plan actually is. They use it in the construction industry, right? Right? Video

I hope you guys understand that this is going to keep getting worse

[AINews] TML-Interaction-Small 276B-A12B latent.space/p/ainews-thinki… Thinking Machines' Native Interaction Models - TML-Interaction-Small 276B-A12B - advances SOTA Realtime Voice and kills standard VAD all our highlights!

I gave it a topic. It came back with a full magazine-style infographic. Charts. Layout. Icons. Colour coding. Dense structured copy. That model is SenseNova U1. And it's open-source. 🧵

Veo 4 vs Seedance 2.0 social media is cooked Veo 4 looks like it will be much better than Seedance 2.0 and it's an omni model which means consistent voice references and image inputs just like Seedance what a time to be alive.. Prompt: A professor writes out a mathematical proof for trigonometric identities on a traditional chalkboard, explaining the step he is currently on in the equation. Video Video

GOOGLE 🔥: An upcoming Gemini Omni video model from Google is expected to be much more advanced in video editing, capable of completing tasks like removing watermarks, replacing objects in the video, and more. It is also likely that Google will release 2 versions of this model, including a Pro variant. And I assume what we see isn't Pro? Anime sample 👀 Video