全部 AI 动态

🛰️ @elonmusk 希望在 2027 年底前获得 10 吉瓦的算力。@JensenHuang 表示，Nvidia 现在将 @SpaceX 与云服务提供商客户并列计算，两家公司的工程师正在设计一款用于太空的服务器： "在我们图表的两个部分中，Elon 过去被归在企业类下。我们之前就是这样报告数字的。现在他属于我们的 CSP……他们规模巨大。 他在使用我们的技术来训练自己的 AI 模型之类的，当然还有部署模型……而 GrokBot 的需求，GrokBot 的消耗量真的是在飙升。然后，当然，他还把它租出去。所以这三件事结合在一起，他这样做相当独特…… 他热爱计算与实体建设之间的交叉领域……而且没有人比他更懂能源。 而我们将一起去探索的一个特别的新市场，就是把数据中心放到太空里。所以我们的两个工程团队正在非常紧密地合作，打造一种新型服务器、一种新型架构，它可能在太空和地球上都非常出色。" —— @JensenHuang，@nvidia 首席执行官，在 @BG2Pod 上与 @altcap 对话

interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled knots. years later, I started openai's first large-scale automated interpretability project, finding about 2% of neurons legible with the primitive tools we had with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works https://already-clean.goodfire.com/

We’re rolling out Gemini 4 Argon in Antigravity to paid Google AI Pro subscribers today. Demand has been incredible and we've scaled capacity deliberately to get here. We’ve rolled out a separate weekly-only Argon quota for Pro on top of the existing Gemini bucket for 3.8 Flash so you have the best of Gemini for all your tasks.

大家好，我叫思定，我在找美国的 software engineer 的工作机会。 我之前在 Alipay/Airbnb/Bytedance ，离职前是团队里某一个方向的 tech lead，带 5 个 eng/intern ，负责过超百万 QPS 的服务。同时做中等规模的技术规划，提供技术指导和带新人。绩效考评连续拿 EE 。 离职后从零到一做了 http://babra.ai，应该是留学咨询行业最好的垂类 AI 产品。4 月上线后，大概 $1000 月营收，15% 转化率。 欢迎所有工作机会🙏🙏🙏 祝各位幸福安康。

Gemini 4 Argon from @Google on ARC-AGI (Verified): - ARC-AGI-3: 97.5%, $4.8K (provider adapter harness), 49.8%, $5.8K (standard harness) - ARC-AGI-2: 91.3%, $0.32/task - ARC-AGI-1: 98.5%, $0.16/task Google DeepMind is the second frontier AI lab to score above 95% on ARC-AGI-3.

Code Review is now generally available for Claude Team and Enterprise plans. Reviews are now faster, catch more of the bugs that matter, and admins can pick from Quick, Standard, or Deep review levels for each repo.

Tinker lets customers turn their expertise into cutting-edge models that they own. One year after launch, we look back at what we’ve shipped and what our users are building. https://thinkingmachines.ai/news/one-year-of-tinker/

interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled knots. years later, I started openai's first large-scale automated interpretability project, finding about 2% of neurons legible with the primitive tools we had with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works https://already-clean.goodfire.com/

i was about to pay for a small Codex subscription just for computer use.. honestly, I didn't expect Grok bot's computer use to be this fast... I've been more and more impressed with it every day lately. for day to day stuff, nothing else I've used comes close right now...great job @bot team..

HOLY SH*T, JENSEN HUANG SAYS GROK BOT USAGE IS SKYROCKETING, AND ELON IS DOING SOMETHING ALMOST NOBODY ELSE DOES JENSEN HUANG SAYS GROK BOT USAGE IS SKYROCKETING, AND ELON IS DOING SOMETHING ALMOST NOBODY ELSE DOES nvidia's ceo, spent one minute on elon and grok bot. what he said: 1. elon trains and runs his own models on nvidia chips 2. grok bot demand is climbing fast, jensen called it super cool 3. elon also rents that compute out to other companies 4. building it, running it and renting it: the only other one jensen names is google 5. his take on elon: nobody knows more about energy start small: give grok bot one research task you keep putting off. save this and read the article below to make money with grok bot ↓

New voices. Better accents. And less "checking..." GPT-Live-1.1 brings these updates to Voice in ChatGPT, Codex, and dots, with new voices in French, German, Hindi, Mexican Spanish, Korean, and Japanese. Rolling out today to all users. Coming soon to the API.

the gpt-3 moment for game dev is here a few months ago i built royale: last agent standing royale drops 11 models into a battle royale to test long term agent planning+actions models have gotten WAY better at game dev, so i asked 3 models to rebuild royale's visual engine 🧵

Congratulations to @StepFun_ai on Step 5 Preview going open weights! RadixArk is proud to help power StepFun's first-party production API with SGLang at scale. Production inference means great token economics under strict latency requirements. SGLang and its cache-aware routing stack now bring that to Step 5 Preview in proven production traffic. Building with Step 5 Preview? Deploy with SGLang today!