Mu Cai

@imhaotian @xuandongzhao @elonmusk I know you can always make impossible things possible by light speed implementation, Haotian! Have a good rest. Hope you all the best!

I left xAI earlier this week. It was a difficult decision. The past two years have been an intense, fun, and deeply rewarding journey, and I accomplished things I could not have imagined two years ago. Thank you @elonmusk for the opportunity and for everything I learned at xAI. Thank you @Guodzh for the trust you placed in me and for all the days and late nights we worked through together. And thank you to the entire Omni / Imagine team: thank you for your trust, and for growing together with me. It has been an honor, and I am incredibly proud of what we achieved together. I feel fortunate to have had the chance to work with all of you. At xAI, everything feels possible. I had the chance to work with and learn from some of the most exceptional people I have ever met. I was able to explore across domains: from pretraining to post-training, from language models to multimodal, from perception to generation. Joining xAI was one of the best decisions I have ever made. @grok imagine is special to me. Building video generation models, where I started with almost zero prior knowledge, from 0 to No.1, as an IC and as a lead, alongside an extraordinary team, and shipping it as a great product used by millions, all within 6 months, at age 28: I feel proud. But now it’s time for me to move on. I’m burnt out, and I know my happiness is no longer maximized in my current state. It is sad to say goodbye, but it is just the right time for a change. Best wishes to the Imagine team, you are absolutely the best, and you deserve the best. I will cherish all our memories for the rest of my life. For now, I’m taking a break and giving myself time to figure out what comes next. Posted from Hawaii.

@sanghyunwoo1219 @amilabs Big congrats Sanghyun!

Building something new. I’m joining @amilabs from day one to work on world models with an amazing founding team. A bigger challenge, a bigger bet, and an exciting road ahead.

Congrats William @shenbokui! The semantics and text rendering looks fantastic!

@shenbokui Congratulations William! The semantics and text rendering look fantastic! We are truly living in the era of unified vision!

Excited to introduce Uni-1, our new multimodal model that *unifies* understanding and generation. TLDR: a team of ~15 researchers is going pound-for-pound with nano banana and gpt image 🧵

@LiangJeff95 🐮

从Meta跑路了，搞一些好玩的东西。一句话就可以生成一分钟的视频。

@thaoshibe I am working on this right now

why all AI-generated websites are purple-blue-green gradient 😭😭😭 i am so tired of seeing all website are purple-ish gradient now 😭 please please we should have VLMs-coding assistant -- i need my coding assistant to be able to seeeeeeeeee 🤡😂🥹🫠

Qwen friends: if any of you want a new home to build great models and contribute to the open models ecosystem, please reach out! Lots of exciting things in the roadmap and so much to build ahead of us

@TongPetersb Congratulations Peter! This is a great work for MM pretraining, especially the finding that with video (pixel) generation loss, there is no harm to text performance!

Train Beyond Language. We bet on the visual world as the critical next step alongside and beyond language modeling. So, we studied building foundation models from scratch with vision. We share our exploration: visual representations, data, world modeling, architecture, and scaling behavior! [1/9]

Still remember the day in last year when I chatted with @JustinLin610 about the future of multimodal and agents, Junyang is such a nice and thoughtful perso! History proves that letting the good people do the do the good things is the easiest way to succeed.

@Xinyu2ML @YouJiacheng 简中信息一哥！

中文发一下今天通义大会的内容吧，感觉是没有转机了 1. 首席hr自称这波调整是扩充更多人才，提供更多资源 2. 阿里是模型公司，qwen是集团的事情，而不只是基模的事情，集团来做大闭环，要快速发展，组织形式没沟通好 3. qwen是集团最重要的事情，希望人才来扩大，必然涉及到阵型变化，无论怎么变化希望大家做好。什么东西都不是没有代价的。用junyang一个人的脑子来处理肯定高效，但站着jingren的角度，需要考虑把zhouhao放在什么位置上比较高效，全过程没有考虑过政治因素（btw昨天高层的说法是，zhouhao比较担心一开始融不进qwen团队，所以主动要求把自己先放在jingren下面，高层就答应了） 4. 我们做的事情很宏大，100多个人肯定不够，需要扩张，很难照顾到每个人的想法 5. 吴妈说中国国情特殊，资源很难大家都满意，道歉没有更早知道资源的问题。说是中国最激进寻求算力的ceo，Qwen是第一优先级&尽了中国CEO最大的努力了。 6. 关于资源被集团卡脖子，吴妈说不知道被卡，心里一直优先级是最高的，问题是信息传递流程的问题 7. jingren说一直资源紧张，在做整体规划，然后说自己也是被架空的。然后说内部阿里云不好用是历史原因 8. 然后下面问junyang能不能回来，首席hr说：不能推上神坛&公司不能接受非理性的要求不计代价来挽留，并问台下那大家觉得自己是什么代价呢

@JustinLin610 Best of luck for your future endeavor, Junyang!

me stepping down. bye my beloved qwen.

@simon_zhai @Yuchenj_UW It already happens! I use a mix of en and zh to prompt coding agent.

@Yuchenj_UW I hope one day when the Chinese open source model catches up, ppl will start coding in Chinese 🤪

I’ve noticed something: When Claude is down, no software engineer says, “Fine, I’ll just write code myself.” They complain, then speed-run to Codex or OpenCode. We’ve lost that ancient skill of manual coding. English is now the only programming language.

Update: position has been filled.

Our team at Google DeepMind is looking for a research intern (Summer 2026)! Multimodal agentic model, unified model (world model). Looking for candidates with multiple first-author papers in top ML conferences and strong engineering skills. Email: caimu_hiring@google.com

@yixin_wan_ @TencentGlobal @leike_lk @wyu_nd Congratulations Yixin!

Thrilled to share that MotionEdit is now accepted to #CVPR 2026! 🥳🥳Thanks to my amazing internship mentors from @TencentGlobal AI Lab, @leike_lk @wyu_nd 🤩🤩 See you in Denver! 🚀

@yueqi_song Congrats Yueqi! This will be pretty useful for the community.

Updates: Excited to share that Agent Data Protocol (ADP) is accepted to ICLR 2026 Oral! 🎉 We also added support for 3 new datasets: SWE-Play, MiniCoder, and Toucan, bringing us to 3M trajectories supported. If you're training agentic LMs, try ADP + tell us what dataset/agent format you want next. PRs & requests welcome. Let's make this the open standard for agent training data 🔥 🚀Original post: x.com/yueqi_song/sta… 📄Read our paper: arxiv.org/abs/2510.24702 🌐Check our project website: agentdataprotocol.com

BREAKING: Gemini 3.1 Pro Preview has landed in #1 on SVG Arena by Design Arena with an ELO of 1421 This 87-point lead the largest winning margin that we've seen a model have on SVG Arena since the arena launch Huge congratulations to the @GoogleDeepMind team!

Gemini 3.1 Pro is here: A smarter model for your most complex tasks. Building on the Gemini 3 series, 3.1 Pro is a step forward in reasoning. It's designed for tasks where a simple answer isn’t enough, taking advanced reasoning and making it useful for your hardest challenges.🧵

we introduce 𝙧𝙚𝙡𝙨𝙞𝙢: 𝙍𝙚𝙡𝙖𝙩𝙞𝙤𝙣𝙖𝙡 𝙑𝙞𝙨𝙪𝙖𝙡 𝙎𝙞𝙢𝙞𝙡𝙖𝙧𝙞𝙩𝙮🍑🌎 . captures image logic/abstraction beyond attribute similarity (CLIP/DINO/LPIPS/dreamsim) . enables logic-based retrieval, analogical image gen, &more! . code&data: thaoshibe.github.io/relsim/ 1/n

@SonglinYang4 @MITEECS @thinkymachines Congrats!

Life update at the end of 2025: I’ve completed my PhD at @MITEECS and joined @thinkymachines to work on LLM archs

@Red_Xiao_ Congrats!

Today marks a moment I'll remember for the rest of my life. When we started Manus, few believed that general AI agents could work. We were told it was too early, too ambitious, too hard. But we kept building. Through the doubts, the setbacks, and the countless nights wondering if we were chasing the impossible. We weren't. This isn't just an acquisition. It's validation that the future we've been building toward is real, and it's arriving faster than anyone expected. But this is not the end. The era of AI that doesn't just talk, but acts, creates, and delivers, is only beginning. And now, we get to build it at a scale we never could have imagined. To everyone who believed in us before it was obvious: thank you. The best is yet to come.