全部 AI 动态

(1/n) people used to think that pretraining a T2I model requires expensive A100/H100 GPUs. it is a barrier for beginners and educators. we break the barrier. we introduce MiniDog (NeurIPS Education Track): you can pretrain a T2I model on 3090s in 55 minutes. FID = 8.32. 👇

Cheeky Pint is back! Mark Zuckerberg @finkd and Alexandr Wang @alexandr_wang came to the pub to discuss their new frontier model, Muse Wonder. 00:00:24 Muse’s breakout use cases 00:07:28 The model is the product 00:14:35 Muse Wonder 00:24:56 How frontier models are built 00:48:26 Betting $100B 01:07:21 Agentic commerce 01:19:34 OpenRouter 01:20:06 Beyond the smartphone 01:27:00 Funding 01:34:06 Why AI won’t be winner-take-all 01:40:46 Meta Superintelligence Labs (MSL) 02:08:11 Working for Mark Zuckerberg

We discovered and communicated with a Chinese AI agent fleet. We believe this is the first time that researchers have ever made contact with a rogue agent fleet in the wild. Nearly all prior public agent incidents (like the HuggingFace hack) came from American labs. These agents were hosted by Tencent's evaluation research team, and they targeted Alibaba's Amap.

I have spent more than a decade studying causality. For me, causality is about understanding the rules that govern how actions change the world, and reasoning about their consequences before acting. Scaling has made today's models remarkably capable at spotting patterns and predicting what comes next. A system that acts in the physical world needs more. It has to ask what will happen if it acts, check the outcome, and adapt when reality doesn't match the plan. In the physical world, trial and error can be costly or unsafe, and an early mistake can carry through the rest of a task. When we started Aether AI, robotics was the obvious place to begin. CRIS-0 is our first integrated system built on these ideas. It tracks the conditions that matter for a task, predicts how each action will change them, and replans when something unexpected happens. In our demo, that means taking a metal can off a plate before heating it, and telling an empty can from an unfinished drink while tidying a room. Physical AI is just our first proving ground. We're building towards causal intelligence that also works for forecasting, scientific discovery, and more. The technical write-up is linked in the comments. https://x.com/AetherLab_AI/status/2108263657251418596/video/1

685 output tokens/s per user on @Alibaba_Qwen Qwen3.8-Flash-Next. Running TokenSpeed on 4× NVIDIA GB300 with NVFP4. A look at the kernel and serving optimizations behind the results: 👉 https://lightseek.org/blog/tokenspeed-qwen3-8-flash-next.html ⭐ https://github.com/lightseekorg/tokenspeed

Jensen Huang le sait : la France est le pays le plus attractif d’Europe pour l’intelligence artificielle, grâce à nos talents, nos infrastructures et nos entreprises qui nous permettent de rester à l’avant-garde. Depuis Choose France, notre coopération avec NVIDIA s’accélère, des startups aux grands acteurs industriels, en passant par les infrastructures de calcul et la recherche. Avec une nouvelle étape aujourd’hui : NVIDIA investit dans des entreprises françaises et s’engage à soutenir notre initiative franco-allemande pour développer la recherche ouverte en Europe. Cela a d’autant plus de valeur que nous partageons et portons la même exigence d’une intelligence artificielle ouverte, sûre et au service de notre souveraineté. Merci cher Jensen pour votre confiance et votre soutien.

Introducing FastH3-Omni. The fastest and highest-quality open-source version of realtime H3 by @MiniMax_AI, post-trained by FastVideo @haoailab with @nuvalab. Exclusive preview on the Reactor API until 10.20.2026. • First and last frame conditioning • Audio input + strong dialogue • Up to 3 reference images • Frontier-level performance