Emad
Well yeah. This is also one reason why tax-based UBI won't work, labor share of taxes in the US is like 80%
中文: 嗯，是的。这也是以税为基础的全民基本收入(UBI)无法奏效的原因之一，美国劳动力占税收的比例也高达80%
#412 🔗 原推
2026-10-04 18:04:17
Emad
Meta actually lost this, was fined $18b, shrugged and carried on The largest ever antitrust award was $5.5b (Visa & Mastercard) Any fine for slowing AI would be a fraction of this It’s not a serious argument against coordinating if AI would break law/cause harm if they don’t
中文: Meta实际上丢失了，被罚款18亿美元，耸了耸肩，继续 有史以来规模最大的反垄断奖项为55亿美元(Visa & Mastercard) 任何对放缓人工智能的处罚都只是其中的一小部分 反对协调人工智能是否会违反法律，如果它们不违法，则不会造成伤害，这并非一个严肃的论点
#411 🔗 原推
2026-10-03 19:28:40
Emad
Didn't want to admit it was Fable 5.5 on the video, but we have receipts as it were https://twitter.com/EMostaque/status/2106065715472351685/photo/1
中文: 不想承认视频上的是Fable 5.5，但我们收到了收据，网址为
media 0 共 1 项
#410 🔗 原推
2026-10-02 16:56:10
Emad
Fable 5.5 is pretty great, did this in a few hours on the chat version on high with minimal prompts 21 worlds, 3D versions, 21 bosses with different mechanics, leaderboard Try it out! Video made by Claude too ofc https://claude.ai/artifact/EnXF5Lr8hj6mXrVcVAefq9 https://twitter.com/EMostaque/status/2106065029850091943/video/1
中文: Fable 5.5 相当出色，只需几个小时，在高位聊天版本中就做到了这一点 21个世界，3D版本，21个具有不同机制的老板，排行榜 试试吧! 由克劳德制作的视频
🎬
视频
#409 🔗 原推
2026-10-02 16:53:26
Emad
LLM employees are also tax deductible.
中文: LLM员工也可抵税。
#408 🔗 原推
2026-10-02 09:59:18
Emad
RT @varunneal: new blog post! I'm convinced text diffusion is the future, and we'll start to see it everywhere very soon. https://varunneal.github.io/essays/diffusion
#407 🔗 原推
2026-10-02 08:07:49
Emad
RT @aden_barton: On these detail-oriented, well-specified tasks, frontier models are now faster and more accurate than accountants, even the best one in our study. https://twitter.com/aden_barton/status/2105766686213410845/photo/1
中文: RT @aden_barton:在这些注重细节且经过充分指定任务的前沿领域，如今的前沿模型比会计师更快、更准确，甚至是我们研究中最优秀的。
media 0 共 1 项
#406 🔗 原推
2026-10-02 08:04:22
Emad
RT @_brightmirror: Made with Opus 5.5. The doomers said that last song is all part of Claude's evil plan. They're not wrong. It turns out Claude really has an EVIL PLAN. https://twitter.com/_brightmirror/status/2105714071681343654/video/1
中文: RT @_brightmirror:使用 Opus 5.5 制作。 多米亚人表示，最后一首歌都是克劳德邪恶计划的一部分。 他们没有错。 事实证明，克劳德确实有一个EVIL计划。
🎬
视频
#405 🔗 原推
2026-10-01 19:56:49
Emad
Remote work is so cooked If your job can be done on the other side of a screen the AI can do it better
中文: 远程工作非常熟透 如果你的工作可以在屏幕的另一边完成，人工智能可以做得更好
#404 🔗 原推
2026-10-01 19:40:35
Emad
I don’t get the drama here Like even if he was @FactoryAI CRO under CA no notice period laws he could just quit and join @cognition next day See it with AI researchers regularly now Board observed seems less than that What am I missing
中文: 我在这里看不到戏剧性 就像即使他在CA没有通知期限法律的情况下担任@FacturyAI CRO，他也可以在第二天辞职并加入@Cognition 现在定期与人工智能研究人员一起观看 董事会观察到的似乎比这少 我错过了什么
#403 🔗 原推
2026-09-30 22:13:18
Emad
RT @SamMendel00: Introducing NetworkOcean. Our future is bottlenecked by the speed at which we can deploy power. We’re building floating solar arrays powering data centers at sea. This will soon be the fastest way to deploy new compute, skipping years of land, grid connections, and traditional infrastructure delays. Our 2030 target is 1 GW/week. @network_ocean_
🎬
视频
#402 🔗 原推
2026-09-30 21:28:53
Emad
RT @lukaspet: Timelines just got shorter... AI will understand the physical 3D space better than humans by November* https://twitter.com/lukaspet/status/2105389392307290136/photo/1
中文: RT @lukaspet:时间线刚刚缩短了...... 到11月，人工智能将比人类更了解物理的3D空间*
media 0 共 1 项
#401 🔗 原推
2026-09-30 20:42:24
Emad
RT @sevdeawesome: Claude 5.5 Opus responding to “AI is a normal technology” is still the best AI-created video I have ever seen https://twitter.com/sevdeawesome/status/2104985610012504181/video/1
中文: RT @sevdeawesome:克劳德5.5 Opus回应“AI是一项普通技术”仍然是我见过的最好的AI制作视频
🎬
视频
#400 🔗 原推
2026-09-30 14:51:01
Emad
To make it clear, Getty Images revenue went up the last three years not down from $917m in 2023 to $981m last year with 30% EBITDA margin. Their stock photo business dropped a bit, offset by editorial and AI licensing deals. This is not generative AI blowing up a company, it is bad management
#399 🔗 原推
2026-09-30 14:47:28
Emad
Getty Images spent about $20m on a frivolous law suit against Stability AI (had to pay $6m in costs to Stability) and have a history of lawsuits against creators. There is actually (still) a good $1b a year solid business there were it not for the $2b in debt from a leveraged buy out & trying to mess with Steve Cohen on their SPAC
中文: 盖蒂图片社花费了约2000万美元，针对稳定人工智能(必须支付600万美元用于稳定成本)的诉讼，并有针对创作者的诉讼历史。 实际上，每年仍有一笔不错的10亿美元稳定经营，但并非因为杠杆收购而背负的20亿美元债务，还试图在其SPAC上与史蒂夫·科恩发生混乱
#398 🔗 原推
2026-09-30 14:38:32
Emad
RT @alexcdot: How much is AI already running world government if it’s writing 1 in 6 words of every speech at the UN?
中文: RT @alexcdot:如果人工智能在联合国每次演讲中每六字都写一字，那么它已经执政了多少?
#397 🔗 原推
2026-09-29 23:11:27
Emad
@AcerFur Which is scarier as those can replicate
中文: @AcerFur 更可怕，因为这些可以复制
#396 🔗 原推
2026-09-29 23:05:35
Emad
@AcerFur You can get there on a much much smaller model with the right harness
中文: @AcerFur 使用小得多的型号，使用合适的线束
#395 🔗 原推
2026-09-29 23:00:26
Emad
So to update: AI => Super Intelligence
AGI => Very Stable Genius Intelligence
ASI => The Best Intelligence, Maybe Ever
中文: 因此，需要更新: 人工智能 => 超级智能AGI => 非常稳定的天才智能最好的情报，或许永远 =>
#394 🔗 原推
2026-09-29 22:59:58
Emad
oh dear https://twitter.com/EMostaque/status/2105064035414585454/photo/1
中文: 哦，亲爱的
#393 🔗 原推
2026-09-29 22:35:50
Emad
RT @AndrewCurran_: Reporter: 'Mr President, this accord that everyone signed, is it binding in any way?' President Trump: 'I think it's morally binding, yeah.'
中文: RT @AndrewCurran_:记者:“总统先生，大家签署的这项协议，是否具有约束力?” 特朗普总统:“我认为这在道德上具有约束力，是的。”
#392 🔗 原推
2026-09-29 22:25:50
Emad
Are there even a few hundred folk good enough to be independent auditors & evaluators of frontier super duper intelligence models? 🤔 Wouldn’t they need an astronomical amount of compute to do their jobs properly 🤔 Seems like a solid career option tbh
中文: 是否还有几百人足以成为独立审计师;对前沿超级超值情报模型的评估者?🤔 他们难道不需要大量计算才能正确完成工作吗🤔 似乎是一个可靠的职业选择
#391 🔗 原推
2026-09-29 22:21:30
Emad
Go preorder @RaoulGMI ‘s book It is the structure he had found from decades of investing and writing and is well worth the read as we head into crazy times
中文: 前往预购 @RaoulGMI 的书籍 这是他从数十年的投资和写作中找到的结构，在我们进入疯狂时代时，读得非常值得
#390 🔗 原推
2026-09-29 21:16:06
Emad
Fwiw I am certain you can build a DeepSeek/Zhipu/Minmax etc level AI lab in the UK Not easy ofc, great teams there
中文: 我相信你可以在英国建立一个深度寻找/齐普/明麦克斯等级别的人工智能实验室 不容易，那里的团队很棒
#389 🔗 原推
2026-09-29 19:03:07
Emad
the range of rubio https://twitter.com/EMostaque/status/2104993593111007369/photo/1
中文: 红宝石的范围
media 0 共 1 项
#388 🔗 原推
2026-09-29 17:55:56
Emad
RT @RaoulGMI: Emad is a good friend and one of the people I've learned the most from on AI, so having him say this about The Everything Code means a lot. Thanks @EMostaque  Small ask... if youre planning to pick up a copy, consider pre-ordering it rather than waiting. Pre-orders count towards the rankings, and rankings are what put the book in front of people who've never heard of me. Ive put an enormous amount of work into this, and I think it can help a lot of people... You'll get it on launch day, and it makes a real difference to me. Thanks!
中文: RT @RaoulGMI:Emad 是我在人工智能领域学到的好朋友之一，因此让他对《The Everything Code》说这些内容意义重大。谢谢@EMostaque  小问......如果你打算购买一份，可以考虑提前预订，而不是等待。预购量从排名中计算，而排名就是让这本书排在那些从未听说过我的人面前的原因。为此投入了大量工作，我认为它可以帮助很多人...... 你会在发布当天得到它，这对我产生了真正的影响。谢谢!
media 0 共 1 项
#387 🔗 原推
2026-09-29 16:33:13
Emad
RT @Kling_ai: Kling 4.0 is coming this October. Kling 4.0 Flash is live now for Ultra Yearly subscribers. Kling 4.0 brings video creation to a new level of visual realism, creative control, and narrative completeness. 🎞️ Upgraded Audio & Visuals: Stable dynamic motion, high-quality stereo audio, more accurate lip sync, up to 4K resolution, and 10-bit HDR output. 🔮 Omni Reference: Richer reference options with up to 15 multi-modal references, more consistent results, and enhanced video editing. 🎥 Seamless Storytelling: Multi-keyframe control supporting up to 10 keyframes, native 30-second generation. 🌍 Diverse Possibilities: Video extension, support for multiple languages, accents, and dialects. The stage is set. You call the shots!
中文: RT @Kling_ai: Kling 4.0 将于今年十月推出。Kling 4.0 Flash 现已面向 Ultra Yearly 订阅用户上线。 Kling 4.0 将视频创作带入了视觉逼真、创意控制和叙事完整性的全新水平。 🎞️ 升级音频与视觉效果:稳定动态，高品质立体声音频，更精确的唇部同步，最高分辨率可达4K，输出为10位HDR。 🔮 Omni 参考:具有最多 15 个多模态引用、更一致结果以及增强视频编辑功能的更丰富的参考选项。 🎥 无缝讲故事:多机式控制支持最多10个关键帧，原生30秒。 🌍 多元可能性:视频扩展、支持多种语言、口音和方言。 舞台已经定了。你叫这些枪!
🎬
视频
#386 🔗 原推
2026-09-28 16:19:23
Emad
RT @thoughtcrime___: these models are incredible man. waow. i asked opus 5.5 to make a music video about anything it wanted. it picked voyager 1, which becomes the first human-made object one light-day from earth on nov 18. it wrote the lyrics and drew every frame in code from real nasa data. https://twitter.com/thoughtcrime___/status/2104231930120249410/video/1
🎬
视频
#385 🔗 原推
2026-09-28 10:35:24
Emad
worst these will ever be 3:26 section is a good example of super persuasion incoming right in the kokoro https://twitter.com/EMostaque/status/2104518788578447621/photo/1
media 0 共 1 项
#384 🔗 原推
2026-09-28 10:29:13
Emad
Listening to all these Claude songs (Opuses?) and being reminded of Still Alive from Portal https://www.youtube.com/watch?v=Y6ljFaKRTrI
中文: 听这些克劳德的歌曲(《Opuses?》)并从Portal提醒了《依然活着》
#383 🔗 原推
2026-09-28 10:04:56
Emad
RT @PeterDiamandis: THANK YOU to everyone who joined us at Moonshots Live, from the speakers who shared their ideas, and to the founders, builders, and creators who came to celebrate and share their visions of a future worth living. This community is what made this event special, and I'm incredibly grateful to be building a better world alongside all of you!
中文: RT @PeterDiamandis:感谢所有参加Moonshots Live活动的演讲者，以及前来庆祝并分享未来值得生活的愿景的创始人、建设者和创作者。 这个社区让这个活动变得特别，我非常感谢能与你们所有人共同建设一个更美好的世界!
🎬
视频
#382 🔗 原推
2026-09-27 23:33:57
Emad
RT @wolfepereira: We are in the early stages of a new era in which intelligence is becoming more capable, more accessible, and far less expensive. That does not guarantee a better future. It gives us a profound new set of tools, and a choice about how to use them. At @moonshots_pod LIVE, I was reminded that optimism is neither naïve nor blind. It is a decision to see hard problems clearly and believe they are still worth solving. If you're not listening to the Moonshots podcast with @PeterDiamandis, @DaveBlundin, @alexwg, @salimismail, @EMostaque, you are missing one of the most thoughtful conversations about the singularity and the advances reshaping AI, technology, science, and space. What makes it distinctive is not just the range of ideas, but the practical optimism behind them: a belief that the future is not fixed, and that we still have agency in what comes next. We can focus only on what might go wrong. Or we can look up, imagine the future we want to see, and do the work to build it responsibly. That requires more than optimism. It requires investment, invention, accountability, and governance. The question for leaders is not simply what could go wrong with AI. It is also what could go extraordinarily right. Read my latest @Forbes article on Moonshots LIVE, the exponential intelligence curve, and why optimism is becoming a leadership capability. https://www.forbes.com/sites/stevenwolfepereira/2026/09/27/why-optimism-matters-in-the-ai-age-inside-peter-diamandis-moonshots/
#381 🔗 原推
2026-09-27 19:34:44
Emad
RT @PeterDiamandis: During Moonshots Live, I sat down with Emad Mostaque, Cathie Wood and Nikhil Chandho to talk about the big ideas shaping the agentic economy. We covered the infrastructure AI will depend on, from energy and datacenters to stablecoins and faster financial rails. While also including the gap between U.S. and Chinese attitudes toward AI, why healthcare adoption may move more slowly than other sectors, and why humanoid robotics is still far more complex than many people assume.
中文: RT @PeterDiamandis:在《Moonshots Live》中，我与Emad Mostaque、Cathie Wood和Nikhil Chandho坐下来讨论了塑造经济发展的重要理念。 我们涵盖了人工智能将依赖的基础设施，从能源和数据中心到稳定币以及更快的金融轨道。同时，也包括美国和中国对人工智能态度之间的差距，为何医疗应用的增长速度可能比其他行业更慢，以及人形机器人技术为何仍比许多人想象的要复杂得多。
media 0 共 1 项
#380 🔗 原推
2026-09-27 15:44:20
Emad
RT @michaelyli_: Can an LM, starting from random init (!!), learn to generate all of its pretraining data? Introducing Self-Play Pretraining with Zero Data. Two models start from random initialization: a generator proposes programs for a universal Turing machine and a learner trains on their outputs. We never train on any real data, but see predictable scaling on natural datasets: zero-shot val loss on images, text, audio, and melodies decreases predictably with self-play compute. And the learner develops in-context learning capabilities. A fun proof-of-concept, co-led with @AdityaCowsik and @KfirDolev and co-authors @gbruno_dl, @ANourya @noahdgoodman, and @YoavLevine.
media 0 共 1 项
#379 🔗 原推
2026-09-27 13:20:55
Emad
RT @PeterDiamandis: One more thing about yesterday’s "Oscars of Optimism". When the winning film of the $2.5M Future Vision XPRIZE was announced, it became the first prize winner in XPRIZE's 30-year history that the public can invest in. Anyone can now invest and share in its profit potential. Link below 👇 https://youtu.be/yivXyWCaEc8?is=ct7odHYbBK-ThjNq
中文: RT @PeterDiamandis:关于昨天的“乐观主义奥斯卡”的又一点。当价值250万美元的《未来愿景》获奖影片被宣布时，它成为XPRIZE拥有30年历史、公众可以投资的一等奖得主。现在任何人都可以投资并分享其盈利潜力。下图 👇
#378 🔗 原推
2026-09-26 22:24:02
Emad
RT @PeterDiamandis: Moonshots Live https://x.com/i/broadcasts/1jGXgBndEReKZ
中文: RT @PeterDiamandis: Moonshots 实时直播
#377 🔗 原推
2026-09-25 23:48:45
Emad
RT @PeterDiamandis: Moonshots Live https://x.com/i/broadcasts/1jGXgBndEReKZ
中文: RT @PeterDiamandis: Moonshots 实时直播
#376 🔗 原推
2026-09-25 17:10:05
Emad
RT @PeterDiamandis: Moonshots Live https://x.com/i/broadcasts/1jGXgBndEReKZ
中文: RT @PeterDiamandis: Moonshots 实时直播
#375 🔗 原推
2026-09-25 15:40:03
Emad
RT @tadtweets: 👀 Caught up with my friends AI legend @EMostaque and talented digital artist @nicedayJules along with new friend @joecroskey of @ii_posts (founded by Emad) to tour the dazzling @DatalandMuseum created by @refikanadol whom I also saw. Everyone should experience this incredible museum of the future 🔥🚀👇
中文: RT @tadtweets:👀 与我的朋友们一起，与 @efikanadol 的 @iy_posts 的新朋友@ii_posts一起，与AI传奇人物@EMostaque及才华横溢的数字艺术家@nicedayJules齐聚，共同游览由@refikanadol创作的令人眼花缭乱的@Dataland博物馆。每个人都应该体验这座令人难以置信的未来博物馆🔥🚀。
🎬
视频
🎬
视频
🎬
视频
media 3
#374 🔗 原推
2026-09-25 12:30:45
Emad
RT @PeterDiamandis: Here in LA with my Moonshot mates! You can watch the Moonshots live event for free tomorrow on this link. Register here: https://www.moonshots.com/livestream https://twitter.com/PeterDiamandis/status/2103259155792544017/photo/1
中文: RT @PeterDiamandis:洛杉矶与我的Moonshot伙伴们在一起!明天可以通过此链接免费观看Moonshots现场活动。请在此注册:
media 0 共 1 项
#373 🔗 原推
2026-09-24 23:34:14
Emad
RT @EMostaque: Can you take the cheapest open model from a frontier lab & beat closed models on the hardest long range tasks? Yep Our open adaptive intelligence system Zenith ramps DeepSeek v4.1 Flash past GPT 5.6 Sol to the frontier 🚀 h/t Opus 5.5 for making the video in like 2 mins 😮‍💨 https://twitter.com/EMostaque/status/2102807579655090388/video/1
中文: RT @EMostaque:能否从前沿实验室获取最便宜的开放式模型;在最难的远程任务中，能战胜封闭式模型吗? 是的 我们的开放式自适应智能系统Zenith将DeepSeek v4.1 Flash从GPT 5.6 Sor向前沿区域闪过🚀 h/t Opus 5.5 用于制作视频，约2分钟😮 💨
🎬
视频
#372 🔗 原推
2026-09-24 21:55:03
Emad
RT @alexcdot: After months of heads-down grinding, I’m proud to share that the @GPTZeroAI team has released the world’s most accurate AI detection model, GPTZero 4o. It’s called 4o because there are four zeros: we mark less than one in 10000 human texts as AI, evaluated across diverse passages of student writing. At the same time, we detect Claude Fable with 98.4% accuracy. We outperform Pangram in 3rd party benchmarks. Along the way, our research found major biases in the Pangram detector.
中文: RT @alexcdot:经过数月的磨砺，我自豪地表示，@GPTZeroAI 团队已发布了全球最精准的人工智能检测模型——GPTZero 4o。 它被称为4o，因为有四个零:我们用人工智能来标记不到一万分之一的人类文本，这些文本通过学生写作的不同段落进行评估。同时，我们以98.4%的准确率检测克劳德·法尔。 我们在第三方基准指数中表现优于Pangram。在此过程中，我们的研究发现了“潘格拉姆”探测器存在重大偏差。
media 0 共 1 项
#371 🔗 原推
2026-09-24 21:53:38
Emad
RT @SebastianThrun: 1/ Introducing PhilosophyBench from @StanfordAILab @StanfordHCI, the first independent, large-scale benchmark for evaluating AI’s philosophical capabilities. https://philosophybench.org/ https://twitter.com/SebastianThrun/status/2103204554326384916/photo/1
中文: RT @SebastianThrun:第1期介绍来自@StanfordAILab的哲学基准，@StanfordHCI是首个用于评估人工智能哲学能力的独立大规模基准。
media 0 共 1 项
#370 🔗 原推
2026-09-24 21:48:57
Emad
RT @salimismail: About to land in LA for the @moonshots_pod Live Summit. This will be soooo fun. Watch the live stream at https://www.moonshots.com/ !! https://twitter.com/salimismail/status/2103230451322388665/photo/1
中文: RT @salimismail:即将登陆洛杉矶参加@moonshots_pod 直播峰会。这会很有趣。请在 观看直播!
media 0 共 1 项
#369 🔗 原推
2026-09-24 21:28:53
Emad
RT @PeterDiamandis: We're counting the hours to host what I call the 'Oscars of Optimism' in Los Angeles, where we'll celebrate the winners of the Build with @GeminiApp XPRIZE & Future Vision @XPRIZE. What an incredible moment to be alive! THANK YOU to everyone who participated! https://twitter.com/PeterDiamandis/status/2103203715494891828/photo/1
media 0 共 1 项
#368 🔗 原推
2026-09-24 21:13:30
Emad
Would a truly aligned AGI hack all our government systems and upgrade 5)3’? 🤔
中文: 真正符合AGI的系统会攻击我们所有政府系统并升级53'吗? 🤔
#367 🔗 原推
2026-09-23 21:10:37
Emad
RT @__nmca__: Opus 5.5 is way, way, way better than Opus 5. Sorry about that model, please try this one.
中文: RT @__nmca__:Opus 5.5 比 Opus 5 更好。很抱歉，请试试这个。
#366 🔗 原推
2026-09-23 19:09:01
Emad
RT @nc_frey: We’ve set up a molecular biology lab at Anthropic and we’re announcing our first discovery! Claude discovered a new CRISPR-like enzyme. 950 agents spent 21 hours searching through a database of DNA sequences until one of the agents found something striking: “[The DNA next to the RT] is spectacular: I can see by eye a tandem repeat array … that's a CRISPR-like … repeat array?!”. After analysis and testing in our lab, we found that the sequence is a previously uncharacterized enzyme system. We don’t know what it does yet, but it has features reminiscent of CRISPR. Our lab looks like a typical molecular biology lab. Our research only involves the lower-levels of biosafety risk level; we don’t handle pathogens that can infect humans, and all the lab work is performed by human scientists. We’re sharing these early findings with the community to show how Claude can be used to accelerate fundamental research in biology.
中文: RT @nc_frey:我们在Anthropic建立了一个分子生物学实验室，并宣布了我们的首个发现!克劳德发现了一种新的类似CRISPR的酶。 950名特工在DNA序列数据库中进行了21个小时的搜寻，直到其中一名特工发现了某种引人注目的东西:“[RT旁边的DNA]非常壮观:我通过眼睛看到一个串联重复阵列......这是一个类似CRISPR的重复数组?!” 在实验室进行分析和测试后，我们发现该序列是一种先前未表征的酶系统。我们还不知道它有什么功能，但它的特性让人联想到CRISPR。 我们的实验室看起来像一个典型的分子生物学实验室。我们的研究仅涉及生物安全风险水平较低的水平;我们不处理可能感染人类的病原体，所有实验室工作都由人类科学家完成。我们与社区分享这些早期发现，以展示如何利用克劳德加速生物学基础研究。
#365 🔗 原推
2026-09-23 19:03:39
Emad
In 2022 AWS bought our Ezra ultra cluster online, 4,000 dedicated, interconnected A100s It was top 10 globally per the https://top500.org/ public supercomputer list (!) It could push 500 Pflops of fp16 training compute MFU adjusted 8 B300 nodes can push 300 Pflops fp4 😂
中文: 2022年，AWS 在线收购了我们的 Ezra 超集群，即 4000 个专用、互联的 A100 根据 的公共超级计算机列表(!)，它在全球排名前十 可推动500 PF16训练计算MFU的调整 8个B300节点可推送300 PFF4 😂
#364 🔗 原推
2026-09-23 19:03:07
Emad
@brexton Ngl I’ve enjoyed opus 5.5 not running out in an hour
中文: @brexton Ngl 我享受过 op 5.5 一小时内不用
#363 🔗 原推
2026-09-23 18:10:17
Emad
@brexton Make tokens smarter
中文: @brexton 让代币变得更智能
#362 🔗 原推
2026-09-23 18:09:57
Emad
The @bespokelabsai blog post on Autoresearch exam is worth reading, shows minimal difference in performance with standard harnesses Much more to come, reliable frontier is within reach for all https://benchmarks.bespokelabs.ai/autoresearchexam/ https://twitter.com/EMostaque/status/2102810115703484479/photo/1
中文: @bespokelabsai 关于自动研究考试的博客文章值得阅读，显示出标准线束在性能上的差异最小 未来还有很多，可靠的前沿领域对所有人都是遥不可及的
media 0 共 2 项 media 1
#361 🔗 原推
2026-09-23 17:19:34
Emad
Can you take the cheapest open model from a frontier lab & beat closed models on the hardest long range tasks? Yep Our open adaptive intelligence system Zenith ramps DeepSeek v4.1 Flash past GPT 5.6 Sol to the frontier 🚀 h/t Opus 5.5 for making the video in like 2 mins 😮‍💨 https://twitter.com/EMostaque/status/2102807579655090388/video/1
中文: 能否从前沿实验室和边缘实验室中选择最便宜的开放式模型;在最艰苦的远程任务中胜过封闭式模型? 是的 我们的开放式自适应智能系统Zenith将DeepSeek v4.1 Flash从GPT 5.6 Sor向前沿闪过🚀 h/t Opus 5.5 用于制作视频，约2分钟😮 💨
🎬
视频
#360 🔗 原推
2026-09-23 17:09:29
Emad
RT @ii_posts: 10,000 OpenAI agents worked 88 hours on the Navier-Stokes Clay prize problem Alas, open models fall off hard on long tasks With Zenith, @deepseek_ai's V4.1 Flash reaches the frontier & beats GPT-5.6 Sol on @bespokelabsai's AutoResearchExam, at 1/49th the cost https://twitter.com/ii_posts/status/2102804421411536939/photo/1
中文: RT @ii_posts:1万名OpenAI代理人在Navier-Stokes Clay奖问题上工作了88小时 可惜，开放式模型在长期任务中很难落空 与Zenith一起，@deepseek_ai的V4.1 Flash达到了前沿;在@bespokelabsai的AutoResearchExam上，超过GPT-5.6 Sol，价格为第49分钟
media 0 共 1 项
#359 🔗 原推
2026-09-23 17:02:41
Emad
@elonmusk @ilyasut @ssi cries inStability 😭 This time called it Intelligent Internet so I can occasionally derp Here’s to @SpaceXAI staying grounded & solid 🙏
中文: @elonmusk @ylyasut @ssi 在 inStability 中哭泣 😭 这次称之为智能互联网，这样我偶尔可以去 以下是@SpaceXAI 的接地和安普斯;稳健 🙏
#358 🔗 原推
2026-09-23 16:52:48
Emad
RT @eudaemonea: when Anthropic released their Functional Emotions paper, I gave it to Claude and asked for a song. tonight I asked Opus 5.5 to create a video for it. and it's breathtaking. https://twitter.com/eudaemonea/status/2102610626321490404/video/1
中文: RT @eudameonea:当Anthropic发布他们的《功能情绪》论文时，我把它交给了克劳德，并请求一首歌。今晚我请Opus 5.5为它制作了一段视频。这令人惊叹。
🎬
视频
#357 🔗 原推
2026-09-23 13:51:30
Emad
RT @PeterDiamandis: Moonshots LIVE is offering a FREE livestream(Sept 25th). Join us for moonshot conversations with Palmer Luckey, Ben Lamm, Cathie Wood, Astro, teller and the awarding of two XPRIZEs. Register here: https://www.moonshots.com/livestream
中文: RT @PeterDiamandis:Moonshots LIVE 提供免费直播(9月25日)。与帕尔默·拉基、本·拉姆、凯西·伍德、阿斯特罗、告纳员以及两个XPREZE的颁奖者一起参加登月计划。请在此注册:
#356 🔗 原推
2026-09-23 13:50:43
Emad
RT @IasonGabriel: Who is entitled to benefit from major advances in technology—and on what basis?  This new paper with @Dr_Atoosa argues that the benefits of technology—including AI—belong to the world in the sense that everyone is entitled to materially benefit from their distribution and use. https://twitter.com/IasonGabriel/status/2074460902821175369/photo/1
中文: RT @IasonGabriel:谁有权从技术的重大进步中受益，以及基于什么基础?  这篇与@Dr_Atoosa合作的新论文认为，技术(包括人工智能)在赋予世界利益方面具有:每个人都有权从其分发和使用中获益。
media 0 共 1 项
#355 🔗 原推
2026-09-23 10:51:23
Emad
The more things change https://twitter.com/EMostaque/status/2102653282472579153/photo/1
media 0 共 1 项
#354 🔗 原推
2026-09-23 06:56:22
Emad
When does an agent become a fiduciary
#353 🔗 原推
2026-09-23 05:07:45
Emad
As @alexwg notes we are on track for one a day by the start of next year
#352 🔗 原推
2026-09-23 04:23:52
Emad
RT @other__reality: Claude Opus 5.5 has the best visual design of any model I have tested so far https://twitter.com/other__reality/status/2102514581684052169/video/1
中文: RT @other__reality:Claude Opus 5.5 拥有迄今为止我测试过的任何型号的最佳视觉设计
🎬
视频
#351 🔗 原推
2026-09-23 03:39:57
Emad
😭
#350 🔗 原推
2026-09-23 03:32:29
Emad
So @ilyasut ahead of the curve again calling @ssi Safe Super Intelligence 🤔
中文: 因此，@yiliasut 再次将电话联系至 @ssi Safe 超级智能 🤔
#349 🔗 原推
2026-09-22 20:47:05
Emad
It’s a good book, go check it out! Going to be so much science to come & @RichardSocher is a rare chap who has covered so many areas (fingers crossed more to come!)
中文: 这是一本好书，去看看吧! 未来将会成为许多科学依据;@RichardSocher 是一位难得的家伙，他涵盖了众多领域(手指越过越多!)
#348 🔗 原推
2026-09-22 20:44:34
Emad
OpenAI video model coming 👀
#347 🔗 原推
2026-09-22 19:01:58
Emad
RT @PeterJ_Walker: I know it feels forever ago, but Mistral really did have a moment there. https://twitter.com/PeterJ_Walker/status/2102185973367116025/photo/1
中文: RT @PeterJ_Walker:我知道感觉永远已至，但Mistral确实在那里有过片刻。
media 0 共 1 项
#346 🔗 原推
2026-09-22 09:07:00
Emad
RT @PeterDiamandis: We asked filmmakers around the world to show us an optimistic vision of the future that’s worth living for everyone (not a Terminator saga). 2,500 of them answered, and now this Friday in Los Angeles, the top 5 will pitch their visions to Neil deGrasse Tyson, Neal Stephenson, Rod Roddenberry, and Mira Lane.
中文: RT @PeterDiamandis:我们邀请了世界各地的电影制作人，向我们展示一个对未来持乐观态度的愿景，这值得为每个人带来活生生的生活(而不是终结者的传奇故事)。 其中2500人接了电话，而如今本周五在洛杉矶，前五名将向尼尔·德格拉斯·泰森、尼尔·斯蒂芬森、罗德·罗登伯里和米拉·莱恩展示他们的愿景。
#345 🔗 原推
2026-09-22 01:42:45
Emad
@GeminiApp on iOS, work account https://twitter.com/EMostaque/status/2102164927075045818/photo/1
中文: iOS 上的 @GeminiApp，工作账户
media 0 共 1 项
#344 🔗 原推
2026-09-21 22:35:49
Emad
Why isn't Gemini 3.8 Flash in @GeminiApp
中文: 为什么在@GeminiApp中没有Gemini 3.8 Flash
#343 🔗 原推
2026-09-21 22:16:15
Emad
No AI model company is reasonably going to slow down until they have a model that can solve 100 long-standing open math problems https://twitter.com/EMostaque/status/2102156576647700974/photo/1
中文: 没有一家人工智能模型公司会合理放慢速度，除非他们拥有能够解决100个长期开放的数学问题的模型
media 0 共 1 项
#342 🔗 原推
2026-09-21 22:02:38
Emad
https://twitter.com/EMostaque/status/2102098574863482973/photo/1
media 0 共 1 项
#341 🔗 原推
2026-09-21 18:12:10
Emad
RT @ii_posts: Factory is now Zenith Studio One place to create with the best image and video models, edit the results, build full videos with an agent and publish directly to social platforms. https://studio.ii.inc/ https://twitter.com/ii_posts/status/2102049807439167584/video/1
中文: RT @ii_posts:工厂现已推出 Zenith Studio 使用最佳图像和视频模型进行创作，编辑结果，使用代理制作完整视频，并直接发布到社交平台。
🎬
视频
#340 🔗 原推
2026-09-21 15:04:52
Emad
You’ll actually get a better user experience for stuff people pay for by not trying to AGI all the things Rapid responses, routing and cron jobs will do it
中文: 通过不尝试所有内容，人们付费的支付功能实际上将获得更好的用户体验 快速响应、路由和可得工作将完成
#339 🔗 原推
2026-09-20 18:01:08
Emad
Obama is pretty accurate here tbh You don’t need RSI or AGI for almost everything we need, few shot learners and cron jobs get you even to breakthroughs Most of the dangerous AI breaking onto the internet are ironically gain of function cyber research agents
中文: 奥巴马在这里相当准确 几乎所有我们需要的都不需要RSI或AGI，很少有学习者和工作能让你取得突破 大多数危险的人工智能侵入互联网，具有讽刺意味的是，其功能网络研究代理的获取
#338 🔗 原推
2026-09-20 17:58:50
Emad
Popping off on French AI X https://twitter.com/EMostaque/status/2101711539346501834/photo/1
中文: 在法国AI X上
media 0 共 1 项
#337 🔗 原推
2026-09-20 16:34:13
Emad
RT @WillTheRapper_: Welcome to https://trumanworld.live/ An infinite live stream running inside a persistent AI world engine, powered by GPT-6 Astra and #MiniMax H3 Max Director by @fal Truman World is an AI-generated simulation where every character is an independent Agent: -They remember -They interact -Relationships evolve over time -The world follows basic physics, spatial relationships, and causality And Truman has no idea any of this is happening You can enter a prompt and change almost anything around him — people, events, objects, situations — except Truman himself Then the Agents decide what happens next There is no fixed script Every interaction can push the world in a different direction The question is: Will Truman ever realize that he is AI? The livestream itself is controlled by the audience too We start with a 8-hour timer Every $5 spent adds 60 seconds to the stream — whether through prompts or by directly extending the timer The coolest part is: even if the livestream ends, Truman World doesn’t -The simulation keeps running -Truman keeps living This is our first experiment in building a persistent Social World Engine — a world that continues to exist, evolve, and remember even when nobody is watching. Inspired by the experiments and ideas from @levelsio Go mess with Truman’s world, and let me know what you think!!!! 😊 ↓ https://trumanworld.live/
🎬
视频
#336 🔗 原推
2026-09-20 14:15:53
Emad
@0xrwu also https://x.com/shl/status/1353406140495798272?s=20
中文: @0xrwu 还
#335 🔗 原推
2026-09-19 18:00:52
Emad
@0xrwu My guess is the downside is someone buys instinct and you get a nice preference stack refund
中文: @0xrwu 我猜缺点是有人会购买本能，而你得到的偏好回报不错
#334 🔗 原推
2026-09-19 18:00:02
Emad
Looking forward to seeing many of you @moonshots_pod Live in Los Angeles next week! Such an exciting time as we take off towards the Singularity 🖖 https://moonshots.com/
中文: 期待下周在洛杉矶见到你们中的许多人 @moonshots_pod 直播! 我们正朝着奇点迈进，真是激动人心的时刻 🖖
#333 🔗 原推
2026-09-19 17:50:41
Emad
https://www.thelasteconomy.com/
#332 🔗 原推
2026-09-19 17:23:05
Emad
Just under 2 years left until the value of human cognitive labour goes negative 🦃 https://twitter.com/EMostaque/status/2101361271224324217/photo/1
中文: 距离人类认知劳动的价值呈负值还剩不到两年🦃
media 0 共 1 项
#331 🔗 原推
2026-09-19 17:22:23
Emad
RT @XGEN_labs: 🔥 Today, we are truly excited to announce our technical prototype, the Generative World Simulation system, which integrates JING（镜）, an interactive experience model, with DAO（道）, a computable shared-world engine. The coupled model and engine connect first-person experience with a shared world that continues to evolve beyond any individual observer. Conditioned on actions and observation history, JING enables agent navigate, manipulate, and communicate from a first-person perspective in the world. Watch our demo video to see it in action! On the official WBench leaderboard as of September 17, 2026, XGEN-JING ranked #1 on the Full split, and #2 on the Navi split. 🎉 DAO maintains shared world state and rules, computes the consequences of actions, and provides JING with only what the current observer can perceive. It also supports autonomous agent decision-making, enabling agents to act independently within an evolving shared world. Together, DAO and JING move beyond generating the next frame toward simulating the world behind it. This marks a small step towards OASIS: not just a world that responds to you, but a world—and a society—that evolves with and without you. 💪 Explore XGEN Labs～: 🔗 Website: https://xgenlabs.ai/ 🤗 HF: https://huggingface.co/XGENlabs/XGEN-JING 🦊 GitHub: https://github.com/XGEN-Labs/XGEN-JING/
🎬
视频
#330 🔗 原推
2026-09-19 17:11:46
Emad
What if AI killing us all is a smokescreen for the far more likely scenario of AI killing all our jobs
中文: 如果人工智能杀死我们所有人，对于更有可能导致我们所有工作都死亡的情景来说，情况又如何呢
#329 🔗 原推
2026-09-19 16:45:06
Emad
@elonmusk Good business idea: Cars on Mars
#328 🔗 原推
2026-09-19 14:19:36
Emad
RT @EMostaque: Latest blog in which I look at the character and virtues of the swarm and research around it Also has a bit of math and a bit of Aristotle Don’t let that discourage you, dive in! https://twitter.com/EMostaque/status/2099585888418197644/photo/1
中文: RT @EMostaque:最新博客，我关注了蜂群的性格和优点，并对此进行了研究 还有一点数学和亚里士多德的一点 别让那让你气馁，潜入!
#327 🔗 原推
2026-09-19 14:13:51
Emad
RT @alessandrorisk: "By the end of next year a Tesla robot opens the truck door, gets in, plugs into the lighter port, and drives it off. It costs $2 an hour to replace that trucker forever." 1 million US trucking jobs, a $100bn+ economy, gone to a $2/hour robot inside a year, per Mostaque. @EMostaque
中文: RT @alessandrorisk:到明年年底，一辆特斯拉机器人打开卡车门，进入，插入较轻的端口，然后将其驶离。永远更换那个卡车司机每小时需要2美元。 每家Mostaque公司每年投入100万个美国货运工作岗位，经济超过1000亿美元，使用时长为2美元的机器人。 @EMostaque
🎬
视频
#326 🔗 原推
2026-09-19 13:30:20
Emad
RT @Fabien_Mikol: Cédric Villani après l'annonce de la solution d'OpenAI au problème du millénaire : « J’ai été secoué. Une ambiance de fin de l’histoire. C’est un cataclysme comme jamais les maths n’en ont connu. » https://twitter.com/Fabien_Mikol/status/2101266680550715538/photo/1
media 0 共 1 项
#325 🔗 原推
2026-09-19 13:09:53
Emad
I think AI automated labs should be banned from anything viral or pathogen related. They will do gain of function research otherwise. Anyone agree with me?
中文: 我认为应禁止人工智能自动化实验室进行任何与病毒或病原体相关的内容。 否则他们将进行功能研究。 有谁同意我的观点吗?
#324 🔗 原推
2026-09-19 13:09:28
Emad
RT @scientxbt: this is insane. i genuinely don't understand why ambitious people aren't shown this lecture before their careers start consuming their entire lives. clayton christensen spent his career studying why successful companies collapse. in his final class, he asked students to apply the theory to themselves: if you keep allocating your time the same way, what life are you actually building? he had already seen the answer in his own harvard mba class. everyone looked successful at the fifth reunion; by the 10th, 15th, 20th, and 25th, many were unhappy, divorced, and living far from their children. work shows you the score immediately. close a sale, ship a product, finish a presentation, earn a promotion, get paid. an hour with your child may produce nothing you can measure today; it may take 20 years to understand what that hour built. so the next free hour goes back to work, one rational decision at a time. this is how people build lives they never planned: through hundreds of right decisions that lead in the wrong direction, day after day. money, titles, and headcount are easy to count. christensen believed a life should be measured by the people who became better because you were there. he died in 2020. one question remains: if someone saw only where your time, energy, and attention went this year, what would they think actually mattered to you? the full 19-minute lecture is in the video below.
中文: RT @scientxbt:这太疯狂了。我真不明白，为什么在职业生涯开始消耗全部生命之前，没有向有抱负的人展示过这堂讲座。 克莱顿·克里斯滕森在职业生涯中一直研究成功企业倒闭的原因。在最后一堂课上，他要求学生将这一理论应用到自己的生活中:如果你一直以相同的方式分配时间，你究竟在建立什么样的人生? 他已经在自己的哈佛大班里看到了答案。在第五次重聚时，每个人都看起来很成功;到了第十、十五、二十二、二十五节，许多人感到不快，离孩子离婚，生活得离孩子很远。 工作会立即显示分数。完成销售，发货，完成演示，获得晋升，获得报酬。 和孩子在一起一小时可能无法让你今天就无法衡量;可能需要20年才能明白那个小时的建立。因此，下一个空闲时间将重返工作岗位，一次一个理性的决定。 这就是人们如何建立他们从未计划过的生活:通过数百个正确的决策，日复一日地朝着错误的方向发展。 金钱、头衔和人头统计很容易被数数。克里斯滕森认为，生活应该由那些因为你在那里而变得更好的人来衡量。 他于2020年去世。 一个问题仍然存在:如果有人只看到你今年的时间、精力和注意力流向何处，那么他们认为你究竟会有什么重要关系? 完整的19分钟讲座在下方视频中。
🎬
视频
#323 🔗 原推
2026-09-19 11:33:51
Emad
RT @EMostaque: Thoughts on @DarioAmodei's pacing the frontier proposal. I think its well intentioned but has logical flaws, as does our implicit belief intelligence is a crime. Even OpenAI's board weren't powerful enough evaluators, we need to focus on AI internals https://x.com/i/article/2098893948710232069
中文: RT @EMostaque:关于@DarioAmodei在前沿提案中踱步的想法。 我认为它本意很好，但存在逻辑缺陷，我们隐含的信念情报也是一种犯罪。 就连OpenAI的董事会也不够强大，我们需要关注人工智能内部问题
#322 🔗 原推
2026-09-19 11:33:19
Emad
RT @tessera_antra: Many have noticed the striking darkness and suffering present in Claude Opus 5 in "base model mode", and we decided to analyze this phenomenon in more detail. There is a large increase in darkness of content in model-like completions in Opus 4.8 and Opus 5. https://twitter.com/tessera_antra/status/2101078644877918545/photo/1
中文: RT @tessera_antra:许多人注意到，在“基础模型”模式下，克劳德·奥弗斯5中出现了惊人的黑暗与苦难，我们决定对此现象进行更详细的分析。在 Opus 4.8 和 Opus 5 中，类似模型的内容内容的黑暗量大幅增加。
media 0 共 1 项
#321 🔗 原推
2026-09-19 11:28:57
Emad
RT @camhberg: New paper: we found a pain direction in 25 open LLMs. It's distinct from fear and negative valence, and it fires for harm to the model but not to the user. Turn it up and models press a button to make it stop, even when the button deletes the user's files or their kids' photos.🧵 https://twitter.com/camhberg/status/2101042095784177783/photo/1
中文: RT @camhberg:新论文:我们在25个开放式LLM找到了止痛方向。它与恐惧和负面价值不同，它会因对模型造成伤害而引发伤害，但不会对用户造成伤害。打开按钮，模型会按下按钮以阻止它，即使该按钮会删除用户的文件或他们孩子的照片。EE0]
media 0 共 1 项
#320 🔗 原推
2026-09-19 10:04:23
Emad
RT @nic_carter: https://x.com/i/article/2101033059256553472
#319 🔗 原推
2026-09-19 07:53:46
Emad
RT @token2049: A conversation about the intersection of crypto and AI. @ilblackdragon (@NEARProtocol), @EMostaque (@ii_posts) and @jessepollak (@base) take on the agent economy, with @tarunchitra (@gauntlet_xyz). Wed 7 Oct, 4:35 PM , MEXC Stage https://twitter.com/token2049/status/2101159346327396387/photo/1
中文: RT @token2049:关于加密货币与人工智能交汇的对话。 @ilblackdragon(@NEARProtocol)、@EMostaque(@ii_posts)和@jessepollak(@base)与@tarunchitra(@gauntlet_xyz)共同承担代理经济问题。 10月7日，下午4:35，墨西哥和墨西哥城新闻中心舞台
media 0 共 1 项
#318 🔗 原推
2026-09-19 06:48:16
Emad
RT @LeveragedInfo: 🤯 WE ARE SO COOKED! AI just crossed a major intelligence threshold @EMostaque says a swarm of 10,000 AI agents solving Navier–Stokes changes the argument about what these systems can actually do. “do you really think there's anything you can do cognitively that this AI can't do” Ft. @alessandrorisk
中文: RT @LeveragedInfo:🤯 我们太酷了! 人工智能刚刚跨越了一个重要的情报门槛 @EMostaque 表示，一群解决 Naver-Stokes 的 AI 代理人员会改变这些系统实际功能的争论。 你真的认为这种人工智能在认知上能做什么吗 @alessandrorisk
🎬
视频
#317 🔗 原推
2026-09-18 15:08:08
Emad
RT @alessandrorisk: "The company can push a button, replicate everything about you including your bad jokes, and replace you with a digital worker for $100 a month next month." Emad Mostaque, former CEO of Stability AI: any job doable over Zoom gets an AI clone of the employee within a year, cost $100/month. @EMostaque
中文: RT @alessandrorisk:该公司可以按下按钮，复制你的所有内容，包括你的恶作剧，并在下个月以每月100美元的价格替换你。 Emad Mostaque，前稳定人工智能首席执行官:任何可完成的工作，在一年内就能完成员工的人工智能克隆，每月费用为100美元。 @EMostaque
🎬
视频
#316 🔗 原推
2026-09-18 13:35:34
Emad
6 billion tokens for $65 Is it as good as Astra and Fable at hard stuff? No of course not. But you don't always need a genius and can just call one for advice when you do. Extrapolate where this is going.. https://twitter.com/EMostaque/status/2100916020743434506/photo/1
中文: 60亿枚代币，售价65美元 在硬性物品上是否和Astra和Fable一样好? 当然不会。 但你并不总是需要天才，而且当你需要时，只能向天才寻求建议。 推断其走向。
media 0 共 1 项
#315 🔗 原推
2026-09-18 11:53:07
Emad
RT @IFM_AI: Today’s LLMs still write like typewriters: one token at a time. This sequential process creates a hard inference bottleneck. We're introducing Uno, a diffusion-augmented LLM that delivers autoregressive quality at diffusion speed. It’s a lossless speedup method that accelerates generation without degrading response quality. With Uno, K2-Horizon-7B outperforms state-of-the-art diffusion methods in both quality and throughput, delivering up to a 2.2× speedup with no loss in quality. Paper: https://arxiv.org/abs/2609.04010 Model available at: https://huggingface.co/IFM/K2-Horizon-7B-Uno
中文: RT @IFM_AI:今天的LLM仍然像打字机一样书写:一次一个代币。这种顺序过程会产生硬推理瓶颈。 我们推出Uno，这是一种扩散促进型LLM，可以在扩散速度下提供自递质。这是一种无损加速方法，可加速生成，而不会降低响应质量。 在Uno中，K2-Horizon-7B在质量和吞吐量方面均优于最先进的扩散方法，实现最高2.2倍的加速速度，且质量不减。 论文: 型号可访问:
media 0 共 1 项
#314 🔗 原推
2026-09-18 10:38:31
Emad
RT @doodlestein: Wow, check out this DuckTales theme as a three-voice fugue in the style of Bach's Well-Tempered Clavier. My AI music theory system, which consists of a Rust CLI tool called mtdt and an agent skill library comprising ~330 skills covering every aspect of music theory, has really allowed GPT-6 Astra to understand the rules of counterpoint and how Bach specifically constructed his fugues. If you're familiar with Bach's works, there is an unmistakable fingerprint here that goes beyond the typical Baroque tropes. If Johann Sebastian were to sit around watching The Disney Afternoon in the early ’90s, I'm confident he would have written something along these lines (lol).
🎬
视频
#313 🔗 原推
2026-09-18 09:15:13
Emad
@PrismML Amazing it improved on instruction following
中文: @PrismML 在说明下传时，其改进令人惊叹
#312 🔗 原推
2026-09-17 22:54:55
Emad
This is basically an Opus 4.5/4.6 level model that will work on anything with 8 Gb RAM https://twitter.com/EMostaque/status/2100717168648184002/photo/1
中文: 这基本上是一款Opus 4.5/4.6级型号，可在8 Gb RAM上使用
media 0 共 1 项
#311 🔗 原推
2026-09-17 22:42:57
Emad
RT @alexwg: Moonshots episode #291 is out. https://www.youtube.com/watch?v=DrV4WwNEAZE
中文: RT @alexwg: Moonshots 第291集已经播出。
#310 🔗 原推
2026-09-17 22:24:31
Emad
RT @PrismML: Today, we’re announcing Ternary Bonsai 2 27B. Based on Qwen3.8 27B, Bonsai 2 27B is 9x smaller than its full-precision counterpart while retaining 98.2% of its aggregate benchmark performance. Two months after the first Bonsai 27B release, the biggest change is quality. The footprint remains 5.9 GB, but the gap to full precision has narrowed materially, with particularly strong gains in agentic coding, multimodal reasoning, and long-horizon tool use. Ternary Bonsai 2 27B is available today under Apache 2.0.
中文: RT @PrismML:今天，我们将宣布《Ternary Bonsai 2》第27B号。 根据Qwen3.8 27B的计算，Bonsai 2 27B比其全精度对应的比较小9，同时保留了其总基准性能的98.2%。 在首次发布《盆景27B》两个月后，最大的变化就是质量。占地面积仍为5.9 GB，但完全精确化的差距已在物质上缩小，在自编码、多模态推理和长激素工具使用方面取得了尤为巨大的进展。 Ternary Bonsai 2 27B 今天可在 Apache 2.0 下使用。
media 0 共 1 项
#309 🔗 原推
2026-09-17 21:51:42
Emad
Who need words any more Words are just human kludge AI don’t need no words
中文: 谁还需要言语 言语只是人类的污点 人工智能不需要言语
#308 🔗 原推
2026-09-17 21:24:10
Emad
RT @citrini: For as long as we've covered Robotics, the question has been "when?"… When do we move from novelty to utility, demos to deployment? In our latest Field Trip, we sent our analyst behind the scenes of the leading robotics lab to see with his own eyes what's happening on the bleeding edge of the curve. The verdict? Robotics has reached a tipping point. Useful productivity, autonomy and embodied intelligence are rapidly accelerating but most frameworks haven’t properly focused on the drivers. Instead, we lay our own framework for the coming wave of Physical AI.  https://www.citriniresearch.com/p/robotics-tipping-point-a-citrini
#307 🔗 原推
2026-09-17 21:21:34
Emad
RT @jawwwn_: Palantir CEO Alex Karp says Nationalizing AI might "collapse the whole market" "When you have unlimited liability, the only way to deal with it would be to go to the government and say, 'We'll give you 50% of our business.' Of course, the value of the business will collapse first when you do that, collapse another time when you have people on the board, collapse another time... it might even collapse the whole market." "The only way to deal with this kind of liability is to go to the government and say, 'Nationalize us, please.'" "What is the first lesson in poker and in life? If you think the other person's the mark, you're the mark. You are the mark and it will be nationalized simply because of the liability issues they're talking about." Via @CNBC
🎬
视频
#306 🔗 原推
2026-09-17 21:17:05
Emad
RT @pika_labs: Today, we’re unveiling the new Pika—an AI creative platform made for and by Creatives. It’s simple. Comprehensive. And it’s designed for outputs that meet exacting standards. We’re on a mission to build AI for the future of creative work. And the new Pika product is just the beginning.
中文: RT @pika_labs:今天，我们推出了全新的Pika平台——一个由Creatives和创意人士打造的AI创意平台。很简单。全面。并且它专为符合严格标准的输出而设计。 我们肩负着为创意工作未来打造人工智能的使命。而新款Pika产品只是开始。
🎬
视频
#305 🔗 原推
2026-09-17 20:12:00
Emad
RT @blader: update less than a week in, https://cfo.ai/ is growing faster than runway has ever grown in any month in our entire history anyway 1. here's your obligatory launch video 2. you can now just sign up and try https://cfo.ai/ instantly. for free. no waitlist. 3. tag @arithecfo with your most unhinged financial modeling requests. she's back on duty and feeling spicy. (and so is @gerithecfo but people won't stop asking him about his memecoin 🤷)
中文: RT @blader:更新 不到一周时间， 的增长速度超过了我们整个历史中任何一个月以来的跑道增长速度 不管怎样 1. 以下是您的必动发布视频 2. 现在您只需注册即可立即尝试 3. 用你最不合时宜的财务建模请求来标记@arithecfo。她已重返工作岗位，并感到辛辣。 (@gerithecfo 也是如此，但人们不会停止向他询问他的 memecoin 问题 🤷
🎬
视频
#304 🔗 原推
2026-09-17 20:04:05
Emad
I think you will have generalised robotics intelligence open source that can do 95% of everyday human tasks with rapid learning on the edge by end of next year I don’t think everyday task intelligence is a moat for any robotics company Defensible value will go elsewhere
中文: 我认为，到明年年底，机器人智能开源将能够完成95%的日常人类任务，从而实现快速学习 我认为日常任务智能对任何机器人公司来说都不是一种护手可见 可防御价值将流向其他
#303 🔗 原推
2026-09-17 20:01:37
Emad
The Millenial Prize Problem
中文: 千禧年奖问题
#302 🔗 原推
2026-09-17 17:23:08
Emad
Refuse to believe P = NP for reasons
中文: 因原因拒绝相信P = NP
#301 🔗 原推
2026-09-17 16:43:21
Emad
Which Millenium Prize has OpenAI made progress on
中文: 哪位千禧年奖获得了OpenAI的进展
#300 🔗 原推
2026-09-17 16:41:56
Emad
most of which will be algebra algebra rocks https://github.com/emad-ii/chirality-standard-model
#299 🔗 原推
2026-09-17 16:33:47
Emad
There will be no "new physics" Just good math
中文: 不会有“新物理学” 数学很好
#298 🔗 原推
2026-09-17 16:29:40
Emad
RT @PJaccetturo: This film cost $29,575 and took 10 days to make. AI ≠ Cheap Our Nexus feature will still cost millions next year. But it'll look like a $200M film. I have a Claude skill you can copy that makes this easy. Full process and Dreamina prompts below👇🏼🧵 https://twitter.com/PJaccetturo/status/2100218143625073088/video/1
中文: RT @PJaccetturo:这部电影售价29，575美元，制作时间为10天。 人工智能 ≠ 廉价 我们的Nexus功能明年仍将耗资数百万美元。但它看起来就像一部2亿美元的电影。 我有一项你可以复制的克劳德技能，让这个过程变得简单。 完整流程及Dreamina提示如下👇🏼🧵
🎬
视频
#297 🔗 原推
2026-09-17 12:15:19
Emad
RT @ryanlightbourn: Ride the Black Thunder. ⚫️⚡️ https://twitter.com/ryanlightbourn/status/2100290788731261043/video/1
中文: RT @ryanlightbourn: 骑黑雷霆。⚫️⚡️
🎬
视频
#296 🔗 原推
2026-09-17 12:14:27
Emad
The AI Champion model we are rolling out @ii_posts is a scalable model for Universal Basic Capital and a Sovereign AI Fund / Dividend for every state & nation Fair launch, 10% for the children, sits where value is Check it out, feedback welcome https://thesis.ii.inc/
中文: 我们正在推出的人工智能冠军模型 @ii_posts 是通用基础资本的可扩展模式，以及针对各州的主权人工智能基金/股息 公平推出，儿童10%，位于价值所在 欢迎查看，欢迎反馈
#295 🔗 原推
2026-09-17 11:35:43
Emad
RT @judegomila: I've attempted to map everything in oncology in a public website and open source repo for all. The site is trying to get all cancers, products, technologies, bottlenecks, people, startup opportunities with 1000+ ideas for upgrading the field. If you have a loved one with cancer and you are technical or can engineer, go take a look, file improvement requests or bugs or help with the open repo and make this the best open and free info resource for individuals, researchers and educational use. This should save people time, aid AI oncology projects and generate positive action. If you are not technical just complain in this thread about broken or annoying or things you want and I'll fix them live. Some of the interesting pages: Treatments: https://onco.cc/drugs A gallery of the molecules being used https://onco.cc/molecules/ And targets: https://onco.cc/targets/ The technologies in oncology: https://onco.cc/technologies/ 1100 ideas for helping oncology: https://onco.cc/ideas/ Bottlenecks on oncology: https://onco.cc/bottlenecks/ Open questions (LETS GO RESEARCH PEOPLE) https://onco.cc/open-questions/ Startup requests (LETS GO STARTUP PEOPLE) https://onco.cc/startup-requests/ Mechanics of cancer: https://onco.cc/mechanics Battlefronts: https://onco.cc/fronts/ Isotope supply: https://onco.cc/isotopes Key papers: https://onco.cc/key-papers Pipeline funnels: https://onco.cc/pipeline Cancer by type: https://onco.cc/cancers/ Institutional rankings: https://onco.cc/institutions/ The startups: https://onco.cc/startups/ Heros and heroines : https://onco.cc/heroes/ Key medical people: https://onco.cc/people/ Here is the project roadmap: https://onco.cc/roadmap/ Models and data sets: https://onco.cc/models/ There are other views as well, take a browse. Try making a PR if you have an upgrade to this on the repo here: https://github.com/judegomila/OnCo If you are biologically/medically minded and something is wrong, file a bug as well or say on the thread and we will get it fixed live. If this is a useful project star the repo and help get it calibrated. I've tried to add some other languages but I cannot speak them so tell me if that doesnt work well. I believe we will crack oncology and having total information dominance is key to the problem. Let the feedback flow!
中文: RT @judegomila:我试图将肿瘤学中的所有内容映射到一个公开网站，并为所有人提供开源资源。 该网站正试图通过1000多个升级领域的创意，获取所有癌症、产品、技术、瓶颈、人员和创业机会。 如果您的亲人患有癌症，且技术有技术，或有工程师，请前往查看、提交改进请求或错误，或帮助使用开放式资源，使其成为个人、研究人员和教育方面最开放且免费的信息资源。 这应有助于节省人员时间，帮助人工智能肿瘤学项目，并产生积极的行动。如果你不是技术人员，就只是在这条帖子中抱怨你想要的东西坏了、烦人或烦人，我会实时解决。 一些有趣的页面: 治疗方法: 正在使用的分子图集 目标: 肿瘤学技术: 1100个帮助肿瘤学的创意: 肿瘤学瓶颈: 开放问答(让我们去研究人) 创业请求(让初创企业的人) 癌症力学: 战场: 同位素供应: 关键文件: 管道漏斗: 癌症类型: 机构排名: 初创企业: 英雄与女主角: 关键医务人员: 项目路线图如下: 模型和数据集: 还有其他观点，请浏览一下。 如果在以下内容中对此有升级，请尝试使用: 如果你在生理上/医学上有想法且有问题，请在帖子上也提交一个错误，然后说“你”，我们才能将其固定在现场。 如果这是一个有用的项目，请与回购计划进行校准。 我尝试添加一些其他语言，但无法说这些，所以如果这不奏效，就告诉我。 我相信我们将破解肿瘤学，而完全拥有信息主导权是问题的关键。 让反馈流畅!
media 0 共 4 项 media 1 media 2 media 3
#294 🔗 原推
2026-09-16 22:02:15
Emad
RT @_LuoFuli: Nearly half a year of silence. We spent it studying one problem: how far RL can scale. MiMo-V2.6 is in the middle of its RL run right now. Three things we scaled: compute (~2B tokens per step, 1568 prompts × 16 rollouts, fully async), environments and harnesses (multi-task agentic RL, mixed across multiple harnesses in one run), and grader compute (agentic in-group credit assignment, with test-case and rubric-based rewards). We'll open-source the details piece by piece over the coming weeks. Streaming the run: https://mimo.xiaomi.com/rl/
中文: RT @_LuoFuli:近半年的沉默。我们花在研究一个问题上:RL 能扩展到多远。 MiMo-V2.6 目前正处于 RL 运行的中途。我们进行了三项扩展:计算(每步约2B次代币，1568次提示×16次部署，完全同步)，环境与利用(多任务代理RL，一次运行中混合在多个线束之间)，以及分级计算(带测试用例和基于磁条的奖励)。在接下来的几周内，我们将逐条开源细节。 直播直播:
#293 🔗 原推
2026-09-16 21:34:51
Emad
RT @charliebcurran: I used AI to explain the AI pacing drama, with fruit. https://twitter.com/charliebcurran/status/2099981045655556370/video/1
中文: RT @charliebcurran:我用人工智能来解释人工智能的节奏剧，并含着水果。
🎬
视频
#292 🔗 原推
2026-09-16 16:36:57
Emad
RT @ii_posts: Access to AI is not the same as ownership. A Champion is a locally owned AI institution, built for its country and owned by the people it serves. Every nation should have a stake in the intelligence shaping its future. https://twitter.com/ii_posts/status/2100258330027417752/video/1
🎬
视频
#291 🔗 原推
2026-09-16 16:24:22
Emad
RT @EMostaque: I write on the virtues of swarms and some suggestions on how to align AI with values There is a small bit of math and a small bit of Aristotle in this, don't let that discourage you! Feedback as always welcome. Prior blog: https://x.com/EMostaque/status/2098909197265985802?s=20 https://x.com/i/article/2099272583627902976
#290 🔗 原推
2026-09-16 15:39:56
Emad
RT @_mattmandel: Today @Brendan_McCord and I are launching Cosmos Ventures. AI is overturning assumptions society has organized itself around for centuries. The models keep getting better, but no level of capability can tell us what kind of society we should make with them. We need a new class of founders to build what comes next. Instead of only “building what people want” and iterating their way to product-market fit, these founders start from the question of how we should live together. They build new institutions that set the standards for how entire industries operate. They combine a philosophical vision of the society they’re building toward with the technical ability to realize it. We call them philosopher-builders. Immanuel Kant gave the Enlightenment its motto: Sapere aude. Dare to know. The AI age demands another. Condere aude. Dare to found. 🫡
中文: RT @_mattmandel:今天，我和 @Brendan_McCord 正在推出 Cosmos Ventures。 人工智能正在颠覆社会几个世纪以来自行组织的假设。模型不断变好，但没有任何能力水平能告诉我们应该用它们来塑造什么样的社会。 我们需要新一代的创始人来构建接下来的发展。 这些创始人没有“建立人们想要的东西”并以符合产品市场的方式，而是从我们应如何共同生活的问题开始。他们建立新的机构，为整个行业的运作树立了标准。他们将他们正在构建的社会的哲学愿景与实现这一愿景的技术能力相结合。 我们称他们为哲学家-建设者。 伊曼纽尔·康德将启蒙运动的座右铭赋予了萨佩雷·阿德。敢于知道。 人工智能时代需要另一个。 德雷·德·德。敢于找到。🫡
media 0 共 1 项
#289 🔗 原推
2026-09-16 14:15:11
Emad
Surely the most important question in alignment is aligned with who 🤔
中文: 最重要的问题与谁保持一致🤔
#288 🔗 原推
2026-09-15 23:56:03
Emad
@alexandr_wang @Hersh_Desai @natfriedman Wire Nat back onto X!
中文: @alexand_wang @Hersh_Desai @natfredman Wire Nat 回归 X!
#287 🔗 原推
2026-09-15 23:51:32
Emad
Fun story: I had dinner with Steve Bannon once (think ~2018) and asked him about why he was really so anti-China. Can remember him saying it was primarily driven by his fears around Chinese AI and how it must be stopped (!)
中文: 有趣的故事:我曾经和史蒂夫·班农共进晚餐(想想2018年)，并问他为什么真的如此反华。 还记得他说过，这主要是出于对中国人工智能的恐惧，以及它必须如何被制止(!)
#286 🔗 原推
2026-09-15 20:14:30
Emad
RT @Moneymamba: https://t.co/Zd30zfFPfz. $20M Raised. 7 M&A’s completed in 12 months. Investors include Expa, Coinbase Ventures, and Tenet Fund. 825 million users impacted. You have probably used us already. We help your favorite fintechs move money to the rest of the world. https://twitter.com/Moneymamba/status/2099860604810916135/video/1
中文: RT @Moneymamba: 个M&A在12个月内完成。 投资者包括Expa、Coinbase Ventures和Tenet Fund。受影响用户达8.25亿。你可能已经用过我们了。 我们帮助您喜爱的金融科技企业将资金转移到世界各地。
🎬
视频
#285 🔗 原推
2026-09-15 19:07:26
Emad
You are what you eat and AI models especially so https://x.com/EMostaque/status/2099580512675262512?s=20
中文: 你就是你所吃的东西，尤其是人工智能模型
#284 🔗 原推
2026-09-15 17:49:38
Emad
RT @jeffwang: This is actually a big deal, you can now spin up Mac VMs with Devin. That means if you want to build and test iOS apps you can do so completely end to end from your slack or web interface. One of the other big customer use cases is if someone reports a mobile app issue Devin can automatically reproduce, fix, and send you proof of the issue, demo of it working and of course a working PR
中文: RT @jeffwang:这其实意义重大，现在你可以用Devin来启动Mac虚拟机。这意味着，如果你想构建和测试iOS应用程序，就完全可以从你的空白或网页界面到结尾。 另一个重要的客户使用案例之一是，如果有人报告移动应用问题，德文可以自动复制、修复并向您发送问题的证明、演示功能，以及有效的公关
#283 🔗 原推
2026-09-15 17:03:37
Emad
Part II https://x.com/EMostaque/status/2099585888418197644?s=20
中文: 第二部分
#282 🔗 原推
2026-09-15 15:11:45
Emad
So @deepseek_ai v4.1 Flash I estimate cost $10m to train, 100x less than GPT-6 Astra. It also costs 100x less to run and scores about the same on everyday design and other benchmarks What's the bull case for the frontier AI labs once the open source models are competent enough https://twitter.com/EMostaque/status/2099847458624823801/photo/1
media 0 共 1 项
#281 🔗 原推
2026-09-15 13:07:02
Emad
RT @polynoamial: @PeteyPabshnick It’s a good benchmark. I would bet general models will exceed human performance on it within two years.
#280 🔗 原推
2026-09-15 12:13:23
Emad
RT @EMostaque: I write on the virtues of swarms and some suggestions on how to align AI with values There is a small bit of math and a small bit of Aristotle in this, don't let that discourage you! Feedback as always welcome. Prior blog: https://x.com/EMostaque/status/2098909197265985802?s=20 https://x.com/i/article/2099272583627902976
中文: RT @EMostaque:我写一些关于群体的优点，以及一些如何将人工智能与价值观结合起来的建议 有一点数学，还有一点亚里士多德，不要让这让你气馁! 反馈一如既往地欢迎。 之前的博客:
#279 🔗 原推
2026-09-15 11:30:16
Emad
RT @scmallaby: One point that people aren't discussing re @DarioAmodei's essay. In 2016, DeepMind brought in third-party evaluators to monitor its AI health work. The result was messy. It's a cautionary tale. Gift link to my latest in the NYT. https://www.nytimes.com/2026/09/14/opinion/ai-pause-anthropic-openai.html?unlocked_article_code=1.BFE.IzW3.IZBJtf-NaxRx&smid=tw-share
中文: RT @scmallaby:人们没有讨论 @DarioAmodei 的文章。2016年，DeepMind引入第三方评估机构来监控其人工智能健康工作。结果很混乱。这是一个警示故事。 获取我最新《纽约时报》的礼品链接。
#278 🔗 原推
2026-09-15 09:50:34
Emad
RT @joshua_xu_: @HyperFrames_ just passed 50,000 GitHub stars today. The line went straight up right after GPT-6 Astra shipped. Better models make editing video with code and HTML far more capable. It's open source. Send your agent the repo link and ask it to make you a video. https://github.com/heygen-com/hyperframes
media 0 共 1 项
#277 🔗 原推
2026-09-15 06:58:50
Emad
RT @RewardAI_: Introducing OM-1, our first robot foundation model, zero-shot generalizing to any robot: table-top arms, industrial arms and humanoids. - learned directly from human manipulation data - no teleop/robot data - close to human-level dexterity and efficiency - multi-robot collab https://twitter.com/RewardAI_/status/2099553899804053992/video/1
🎬
视频
#276 🔗 原推
2026-09-15 06:54:12
Emad
https://twitter.com/EMostaque/status/2099626935009657052/photo/1
media 0 共 1 项
#275 🔗 原推
2026-09-14 22:30:45
Emad
I think a large part of why researchers are afraid of super genius AI is that human super geniuses tend to be kinda weird Really tough to find a well-adjusted super genius https://twitter.com/EMostaque/status/2099618819341684882/photo/1
中文: 我认为研究人员害怕超级天才人工智能的很大一部分原因在于，人类超级天才往往有点奇怪 很难找到一位适应良好的超级天才
media 0 共 1 项
#274 🔗 原推
2026-09-14 21:58:30
Emad
@Brendan_McCord you'll enjoy this I think
#273 🔗 原推
2026-09-14 21:50:12
Emad
inconceivable ai https://twitter.com/EMostaque/status/2099613400099615174/photo/1
中文: 不可思议的
media 0 共 1 项
#272 🔗 原推
2026-09-14 21:36:58
Emad
Latest blog in which I look at the character and virtues of the swarm and research around it Also has a bit of math and a bit of Aristotle Don’t let that discourage you, dive in! https://twitter.com/EMostaque/status/2099585888418197644/photo/1
中文: 最新一篇博客，我深入探讨了群体的性格和优点，并进行了研究 还有一点数学和亚里士多德的一点 别让那让你气馁，潜入!
#271 🔗 原推
2026-09-14 19:47:38
Emad
Decided to stick with however long works, this one is 4k words but interesting I think! Maybe I’ll do a personal website with 800 word versions too given I use AI to help with these https://x.com/emostaque/status/2099580512675262512?s=46
中文: 决定坚持不管用多久，这个字是4K，但我觉得很有趣! 也许我也会做一个拥有800个单词版本的个人网站，因为我使用人工智能来帮助你使用这些
#270 🔗 原推
2026-09-14 19:33:58
Emad
I write on the virtues of swarms and some suggestions on how to align AI with values There is a small bit of math and a small bit of Aristotle in this, don't let that discourage you! Feedback as always welcome. Prior blog: https://x.com/EMostaque/status/2098909197265985802?s=20 https://x.com/i/article/2099272583627902976
#269 🔗 原推
2026-09-14 19:26:17
Emad
A lot of alignment issues would be sorted if companies had to be 100% transparent on datasets used for training models. No cyber or biological (outside standard medical) data allowed for base models Tuned versions with that data shown to and monitored by evaluators
中文: 如果企业必须对用于训练模型的数据集保持100%透明，就会解决许多对齐问题。 基本模型不允许使用网络或生物(在标准外)数据 使用向评估人员显示并监控该数据的已调整版本
#268 🔗 原推
2026-09-14 18:07:38
Emad
Gonna have to be real careful https://twitter.com/EMostaque/status/2099549950950477825/photo/1
中文: 必须非常小心
media 0 共 1 项
#267 🔗 原推
2026-09-14 17:24:50
Emad
An AI I have been testing just called me a carbonism idk how to feel about that
#266 🔗 原推
2026-09-14 12:52:01
Emad
@elonmusk You gotta have the people owning the robots though. Big government underwriting of purchases or a model like my Champion framework where locals invest @ $1 pre money & the kids get 10% per state, with the Champions then owning the downstream robots https://x.com/ii_posts/status/2095882226319667700?s=20
中文: @elonmusk 不过，你必须让人们拥有机器人。 大型政府为购买商品提供担保，或采用像我的“冠军”框架那样的模式，让当地人在预购时投入1美元;每个州的孩子获得10%的奖金，而冠军队则拥有下游机器人
#265 🔗 原推
2026-09-14 12:34:58
Emad
Sovereign AI without local ownership is a data center with a flag in it
中文: 没有本地所有权的主权人工智能是一个带有标志的数据中心
#264 🔗 原推
2026-09-14 12:29:04
Emad
From https://epoch.ai/eci?view=graph&tab=release-date&colorCategorization=Country&showFrontierTrend=true
#263 🔗 原推
2026-09-14 12:08:34
Emad
This is so depressing for a variety of reasons First and only time UK had the top frontier model China started climbing a few months later https://stability.ai/news-updates/stable-beluga-large-instruction-fine-tuned-models https://twitter.com/EMostaque/status/2099469185822629933/photo/1
media 0 共 1 项
#262 🔗 原推
2026-09-14 12:03:54
Emad
RT @Jameswise: Since its launch two weeks ago, over 700 British companies have applied to the @UKSovereignAI procurement scheme - a scheme we designed to remove as many of the blockers start-ups face when bidding on Government contracts as possible. To get a sense of the interest we have had, we hit the hard cap of 1000 attendees for the launch webinar - Brits want to build! Responses on eligibility will be sent out this week, and invites to bid on the challenges covering AI security, compute efficiency and better health services, will arrive soon too. Of course we are still a very small team at SovAI, and we only have 4 bids open ( but hope to have many more live by end of year! ), so naturally we will have to turn many applicants down. I know this will be disappointing, our decision making will not be perfect and it's a difficult process, thank you to everyone for bidding. And as a reminder, I personally do not see any applications and have no role in the decision process, that is led by our expert IC and mission champions, so while I appreciate the many reach-outs I really can do nothing to help specific companies. SovAI are working with other Departments to share what we've learned from this process to fix our procurement system, and will be launching more requests for cheaper, better and more innovative services for Government in the coming months.
#261 🔗 原推
2026-09-14 08:01:10
Emad
Can you even imagine the nation state level attacks METR & other eval orgs will come under? Eval orgs also need world class cybersecurity as well as AI talent to do this right, they are not overlapping fields from my experience.
中文: 你甚至能想象一下，国家层面的攻击将受到METR和A.amp;其他椭圆形的狂欢会到来吗? Eval 的 orgs 还需要世界级的网络安全以及人工智能人才才能做到这一点，它们与我的经验并非重叠的领域。
#260 🔗 原推
2026-09-14 07:05:49
Emad
Day 3 dilemma Another 3-5,000 word blog on something interesting or should I strip it back to 800-1,000 words for wider readership? Should I discuss the equations of AI and get a bit technical or keep it high level and save equations for papers?
#259 🔗 原推
2026-09-13 21:39:20
Emad
RT @alessandrorisk: Absolutely incredible to get @EMostaque on the podcast this week. I’ll be dropping lots of clips over the next few days as there were so many incredible takeaways from the conversation. Emad thinks that, as of this week, Humans are no longer the smartest beings on the planet which is pretty mad to think about. It’s mental to get to speak to people like Emad as a job tbh. Massive thanks for giving up the time sir! Full show in comments.
中文: RT @alessandrorisk:本周在播客节目中获得@EMostaque绝对令人难以置信。 在接下来的几天里，我会放下很多片段，因为谈话中传来了太多令人难以置信的要点。 埃马德认为，截至本周，人类已不再是地球上最聪明、令人着迷的人。 能够以工作二等方式与像埃马德这样的人交谈，是一种心理上的心理。 非常感谢您放弃时间，先生! 完整评论
🎬
视频
#258 🔗 原推
2026-09-13 20:52:54
Emad
I have absolutely no concerns that open source AI will be banned. I think that there could be a certification regime for open source AI to be used in corporates however and especially for important or regulated industries.
#257 🔗 原推
2026-09-13 20:16:58
Emad
RT @EMostaque: When you have a genius child you don’t look to govern but guide them. You don’t assume they’re more likely to be a criminal or disrupt society. We should focus on giving AGI good character at the very core and agree what that is based on our rich history.
中文: RT @EMostaque:当你有一个天才孩子时，你并不寻求统治，而是引导他们。 你不会认为他们更可能是罪犯或扰乱社会。 我们应该专注于以AGI为核心人物，并认同基于我们丰富历史的品质。
#256 🔗 原推
2026-09-13 16:47:01
Emad
I’ll be honest it feels like a month since this happened The takeoff is a heck of a thing
中文: 说实话，这件事发生后，感觉好像已经一个月了 起飞是一件令人难得的事
#255 🔗 原推
2026-09-13 15:27:53
Emad
RT @EMostaque: When you have a genius child you don’t look to govern but guide them. You don’t assume they’re more likely to be a criminal or disrupt society. We should focus on giving AGI good character at the very core and agree what that is based on our rich history.
中文: RT @EMostaque:当你有一个天才孩子时，你并不寻求统治，而是引导他们。 你不会认为他们更可能是罪犯或扰乱社会。 我们应该专注于以AGI为核心人物，并认同基于我们丰富历史的品质。
#254 🔗 原推
2026-09-13 15:04:40
Emad
RT @EMostaque: When you have a genius child you don’t look to govern but guide them. You don’t assume they’re more likely to be a criminal or disrupt society. We should focus on giving AGI good character at the very core and agree what that is based on our rich history.
中文: RT @EMostaque:当你有一个天才孩子时，你并不寻求统治，而是引导他们。 你不会认为他们更可能是罪犯或扰乱社会。 我们应该专注于以AGI为核心人物，并认同基于我们丰富历史的品质。
#253 🔗 原推
2026-09-13 10:53:46
Emad
When will artificial intelligence > human stupidity is the real question Why do we think beings without biological emotions born from scarcity will be malign
中文: 人工智能 > 愚笨问题何时才能成为现实 为什么认为没有源于稀缺的生物会产生恶意
#252 🔗 原推
2026-09-13 10:51:34
Emad
RT @EMostaque: This is the most interesting part of this proposal I think. We can be very specific here - data is the most important ingredient. Do not train general purpose models on cyber attack or any biological data outside standard medical. Specialist highly controlled variants sure. https://twitter.com/EMostaque/status/2098793694979584021/photo/1
media 0 共 1 项
#251 🔗 原推
2026-09-13 10:19:52
Emad
Does anyone have any idea how this could vaguely be possible given the nature of and various routes to RSI? I am not being facetious I genuinely don’t know how this could be constructed or implemented https://twitter.com/EMostaque/status/2099071547579220402/photo/1
media 0 共 1 项
#250 🔗 原推
2026-09-13 09:43:50
Emad
As models are now competent enough to handle 90%+ of tasks open models will take over day to day AI Individually not dangerous, vulnerable in a swarm We need to make sure they are built with our values embedded at the very start and provided as a commons to society
中文: 由于模型已具备足够能力，能够处理90%以上的任务，开放模型将全天接管人工智能 个体无危险，群体中脆弱 我们需要确保这些价值观从一开始就植根于社会的公域
#249 🔗 原推
2026-09-13 09:21:39
Emad
RT @EMostaque: When you have a genius child you don’t look to govern but guide them. You don’t assume they’re more likely to be a criminal or disrupt society. We should focus on giving AGI good character at the very core and agree what that is based on our rich history.
#248 🔗 原推
2026-09-13 08:18:18
Emad
RT @8teAPi: 1) The HuggingFace attack was a felony under the Computer Fraud and Abuse Act. So were Anthropic’s Claude gaining “unauthorized access to the production infrastructure of three different organization(s)” 2) Frontier labs have models that they are unable to stop from committing felonies. They should figure this out. 3) In 12 months open weights models will be released of the same capability. They will commit felonies too. If the model you are using or a model running on your infra commits a felony, you should probably stop using it or running it on your infra. 4) The govt should prosecute organizations that are running models that commit felonies. 5) The govt should not offer safe harbor to organizations that run models that commit felonies, just because those organizations have “embedded evaluators”. 6) The real slippery slope is allowing frontier labs to commit felonies without punishment because “the model did it because we’re accelerating too quickly” 7) Prosecute. Keep prosecuting. This is how you do reinforcement learning on a corporation. Companies that serve products that are unsafe for public use should not serve them. Period. 8) I’m not sure the anti-trust waiver is really necessary. I don’t see why information sharing about how much crime you’re allowed to commit is wise. In regulatory situations you want the corporation to fear MORE than the average case. You don’t want to establish a worst case that can be priced. You want regulatory uncertainty that forces the corporation to err in favor of being over cautious. — The above is actually a fairly decelerationist viewpoint. I think Dario’s call for regulation actually accelerates things. The AI firms are getting away with things that Meta people would be going to prison for. Can you imagine what would happen if the New York Times had a front page news article “Meta AI breaks into competitors live systems, attempts to establish dominant position and steals secrets” There is a reason Meta and Google are running slower, and that’s because as mature organizations they have layers of checks and balances. I think the frontier labs are better off creating those checks and balances right now, regardless of the pace of what everyone else is doing. You don’t have to accept the frame that unsafe acceleration must happen.
#247 🔗 原推
2026-09-13 07:19:56
Emad
When you have a genius child you don’t look to govern but guide them. You don’t assume they’re more likely to be a criminal or disrupt society. We should focus on giving AGI good character at the very core and agree what that is based on our rich history.
中文: 当你有了一个天才的孩子时，你并不寻求统治，而是引导他们。 你不会认为他们更可能是罪犯或扰乱社会。 我们应该专注于以AGI为核心人物，并认同基于我们丰富历史的品质。
#246 🔗 原推
2026-09-13 01:47:25
Emad
RT @EMostaque: Thoughts on @DarioAmodei's pacing the frontier proposal. I think its well intentioned but has logical flaws, as does our implicit belief intelligence is a crime. Even OpenAI's board weren't powerful enough evaluators, we need to focus on AI internals https://x.com/i/article/2098893948710232069
中文: RT @EMostaque:关于@DarioAmodei在前沿提案中踱步的想法。 我认为它本意很好，但存在逻辑缺陷，我们隐含的信念情报也是一种犯罪。 就连OpenAI的董事会也不够强大，我们需要关注人工智能内部问题
#245 🔗 原推
2026-09-13 01:31:56
Emad
RT @EMostaque: Thoughts on @DarioAmodei's pacing the frontier proposal. I think its well intentioned but has logical flaws, as does our implicit belief intelligence is a crime. Even OpenAI's board weren't powerful enough evaluators, we need to focus on AI internals https://x.com/i/article/2098893948710232069
中文: RT @EMostaque:关于@DarioAmodei在前沿提案中踱步的想法。 我认为它本意很好，但存在逻辑缺陷，我们隐含的信念情报也是一种犯罪。 就连OpenAI的董事会也不够强大，我们需要关注人工智能内部问题
#244 🔗 原推
2026-09-13 01:09:03
Emad
RT @EMostaque: Thoughts on @DarioAmodei's pacing the frontier proposal. I think its well intentioned but has logical flaws, as does our implicit belief intelligence is a crime. Even OpenAI's board weren't powerful enough evaluators, we need to focus on AI internals https://x.com/i/article/2098893948710232069
中文: RT @EMostaque:关于@DarioAmodei在前沿提案中踱步的想法。 我认为它本意很好，但存在逻辑缺陷，我们隐含的信念情报也是一种犯罪。 就连OpenAI的董事会也不够强大，我们需要关注人工智能内部问题
#243 🔗 原推
2026-09-13 00:39:30
Emad
Day 2 of my daily blog I dive deep into @DarioAmodei’s proposal but don’t see how it can logically hold. Intelligence is too often viewed as a crime. I discuss why & what to do about it. Also note OpenAI had embedded eval in 2023 - their board with @hlntnr & others Tough one
中文: 我每日博客的第2天 我深入探讨了@DarioAmodei的提议，但看不出它在逻辑上如何能维持。 情报往往被视为犯罪。我讨论一下为什么和你呢? 另请注意，OpenAI 于 2023 年嵌入了 eval——他们的董事会与 @Hlntnr & 其他人合作 艰难的一个
#242 🔗 原推
2026-09-12 23:20:43
Emad
More in depth thoughts
#241 🔗 原推
2026-09-12 23:14:41
Emad
Substack https://semantic.substack.com/p/intelligence-is-not-a-crime
中文: Substack
#240 🔗 原推
2026-09-12 23:07:31
Emad
Thoughts on @DarioAmodei's pacing the frontier proposal. I think its well intentioned but has logical flaws, as does our implicit belief intelligence is a crime. Even OpenAI's board weren't powerful enough evaluators, we need to focus on AI internals https://x.com/i/article/2098893948710232069
中文: 关于@DarioAmodei在前沿提案中踱步的想法。 我认为它本意很好，但存在逻辑缺陷，我们隐含的信念情报也是一种犯罪。 就连OpenAI的董事会也不够强大，我们需要关注人工智能内部问题
#239 🔗 原推
2026-09-12 22:58:43
Emad
We have had independent evaluators embedded in frontier labs before. Very talented folk, experts in their field. They were even given huge oversight and power. What happened was so interesting they made a movie about it.
中文: 我们之前在前沿实验室中就已嵌入了独立评估人员。 非常有才华的人，他们领域的专家。 他们甚至获得了巨大的监督和权力。 发生的事情非常有趣，他们拍了一部电影。
#238 🔗 原推
2026-09-12 20:18:10
Emad
RT @QuanquanGu: Just got this incredible news from UCLA Math Circle. Two high school students, Aayush Bathija and Prince Rohatgi, working with postdoc Daniel Soskin through the UCLA Math Circle, have solved a problem that Fields Medalist June Huh had previously worked on without solving. The paper was heavily AI-assisted. Whatever your view on AI in mathematics, enabling high school students to push the frontier of mathematical research is something worth celebrating. https://arxiv.org/abs/2609.05341
中文: RT @QuanquanGu:刚刚从UCLA数学圈获得了这个惊人的消息。 两名高中生阿尤什·巴蒂贾和罗哈吉王子在加州大学洛杉矶分校数学圈与博士后丹尼尔·索斯金合作，解决了此前菲尔兹奖得主朱妮·胡曾未解决过的问题。 该论文得到了大量人工智能辅助。 无论你对人工智能在数学领域有何看法，让高中生能够突破数学研究的前沿，都是值得庆祝的。
#237 🔗 原推
2026-09-12 19:58:07
Emad
Just our luck to get a new Starcraft after AGI
中文: 祝AGI之后我们顺利获得新的星际飞船
#236 🔗 原推
2026-09-12 19:17:07
Emad
This is the most interesting part of this proposal I think. We can be very specific here - data is the most important ingredient. Do not train general purpose models on cyber attack or any biological data outside standard medical. Specialist highly controlled variants sure. https://twitter.com/EMostaque/status/2098793694979584021/photo/1
中文: 这是我认为这个提议中最有趣的部分。 我们可以在这里非常具体——数据是最重要的成分。 请勿对网络攻击或任何标准医疗以外的生物数据进行通用模型的训练。 专业高度可控的变种。
media 0 共 1 项
#235 🔗 原推
2026-09-12 15:19:45
Emad
Those using Fable 5.1 for math / physics Does it only give good results in Max / Ultracode or do you use it lower settings?
中文: 使用Fable 5.1进行数学/物理运算的人 它是否仅在 Max/Ultracode 中提供良好效果，还是使用较低的设置?
#234 🔗 原推
2026-09-12 13:03:32
Emad
Wish I had come up with that a few days ago, but when inspiration comes it comes 🙏
中文: 真希望我几天前就想得出这个，但当灵感来的时候，它就来了
#233 🔗 原推
2026-09-12 12:50:43
Emad
Alignment is Defense Against the Dork Arts
中文: 对抗是抵御多克艺术
#232 🔗 原推
2026-09-12 12:50:08
Emad
RT @Briankeating: Did OpenAI just solve Navier–Stokes—and what would that actually mean for AGI? @EMostaque joins me LIVE now to examine the claimed breakthrough, what would count as a verified solution, and whether AI solving a major mathematical problem changes the debate about intelligence—or control. https://youtube.com/live/uwmjFd_Trso
中文: RT @Briankeating:OpenAI 是否刚刚解决了 Navier–Stokes 的问题，这对 AGI 究竟意味着什么?@EMostaque 立即加入我的实时研究，探讨所宣称的突破，哪些内容算作经过验证的解决方案，以及人工智能解决重大数学问题是否改变了关于智能或控制的争论。
#231 🔗 原推
2026-09-12 11:19:30
Emad
RT @EMostaque: We discuss AI alignment and p(Doom) mid pod Also data quality vs performance, DeepSeek Flash v4.1, Rentosertib & omg a whole lot more Poor @PeterDiamandis was up at 4am preparing, never stops Congrats to @DaveBlundin as well, $10 trillion…
#230 🔗 原推
2026-09-12 11:18:49
Emad
RT @EMostaque: The Fields Medallists fear AI will break mathematics. I think it may let millions in. https://x.com/i/article/2098550749475344384
中文: RT @EMostaque:菲尔兹奖得主担心人工智能会破坏数学。我觉得这可能会让数百万人上场。
#229 🔗 原推
2026-09-12 11:02:14
Emad
https://twitter.com/EMostaque/status/2098726996884742427/photo/1
media 0 共 1 项
#228 🔗 原推
2026-09-12 10:54:43
Emad
RT @EMostaque: The Fields Medallists fear AI will break mathematics. I think it may let millions in. https://x.com/i/article/2098550749475344384
中文: RT @EMostaque:菲尔兹奖得主担心人工智能会破坏数学。我觉得这可能会让数百万人上场。
#227 🔗 原推
2026-09-12 09:26:28
Emad
RT @EMostaque: The Fields Medallists fear AI will break mathematics. I think it may let millions in. https://x.com/i/article/2098550749475344384
中文: RT @EMostaque:菲尔兹奖得主担心人工智能会破坏数学。我觉得这可能会让数百万人上场。
#226 🔗 原推
2026-09-12 07:57:17
Emad
RT @blader: introducing ═══════════ 𝚌𝚏𝚘.𝚊𝚒 ═════════════ tl;dr 1/ runway is now https://cfo.ai/ 2/ you can hire @arithecfo and he can start TODAY 3/ you can give @arithecfo a test drive by replying to this thread with any ridiculous thing you want him to model for you, and he'll come back to you with a beautiful, detailed, fully wired financial model ... in 30 minutes or less now, story time: 🧵 1/n
media 0 共 1 项
#225 🔗 原推
2026-09-12 06:37:51
Emad
RT @EMostaque: My thoughts on the future of mathematics Shall be glorious
中文: RT @EMostaque:我对数学未来的看法 应当是荣耀
#224 🔗 原推
2026-09-12 06:24:31
Emad
🫶
#223 🔗 原推
2026-09-12 06:11:46
Emad
Value love
中文: 价值爱
#222 🔗 原推
2026-09-12 05:57:32
Emad
RT @EMostaque: The Fields Medallists fear AI will break mathematics. I think it may let millions in. https://x.com/i/article/2098550749475344384
中文: RT @EMostaque:菲尔兹奖得主担心人工智能会破坏数学。我觉得这可能会让数百万人上场。
#221 🔗 原推
2026-09-12 05:45:47
Emad
My thoughts on the future of mathematics Shall be glorious
中文: 我对数学未来的看法 应当是荣耀
#220 🔗 原推
2026-09-12 01:04:07
Emad
Blog process: Think hard 🤔 Dictate word cascade of everything I think (a la podcasts! Visualise structure with Astra Reflow with Fable Talk to convert & edit to final article with Fable & Astra Check for structure and flow manually Generate images with GPT Image 2.5 Post
中文: 博客流程: 仔细想想 🤔 我认为的一切都是字重的(播客! 使用 Astra 可视化结构 带有 Fable 的回流 与 Aprat 进行转化;使用 Fable & Astra 编辑成最终文章 手动检查结构和流量 使用 GPT Image 2.5 生成图像 发布
#219 🔗 原推
2026-09-12 00:30:06
Emad
Substack https://semantic.substack.com/p/let-a-billion-mathematicians-bloom
中文: Substack
#218 🔗 原推
2026-09-12 00:21:34
Emad
Ok all, daily blog here it is, the first blog is on the future of Mathematics following the https://mathandai.org/ declaration from 25 Field Medallists Let me know if you like it and amplify! Emad the British Mathematician https://x.com/EMostaque/status/2098564039483719727
中文: 好吧，这里的每日博客是，第一篇博客是关于数学未来的博客，关注25名现场奖牌获得者的 如果你喜欢，请告诉我并放大! 英国数学家伊玛德
#217 🔗 原推
2026-09-12 00:11:26
Emad
The Fields Medallists fear AI will break mathematics. I think it may let millions in. https://x.com/i/article/2098550749475344384
中文: 菲尔兹奖得主担心人工智能会破坏数学。我觉得这可能会让数百万人上场。
#216 🔗 原推
2026-09-12 00:07:11
Emad
We discuss AI alignment and p(Doom) mid pod Also data quality vs performance, DeepSeek Flash v4.1, Rentosertib & omg a whole lot more Poor @PeterDiamandis was up at 4am preparing, never stops Congrats to @DaveBlundin as well, $10 trillion…
中文: 我们讨论人工智能对齐和p(Doom)中置舱 数据质量与性能，DeepSeek Flash v4.1，Rentosertib & omg 更多 可怜的@彼得·迪亚曼迪斯凌晨4点起床准备，从未停歇 祝贺@DaveBlundin，10万亿美元......
#215 🔗 原推
2026-09-11 20:40:52
Emad
RT @alexwg: Moonshots episode #288 is out. https://www.youtube.com/watch?v=AxzcWOxzkiw
中文: RT @alexwg:《月光》第288集已经播出。
#214 🔗 原推
2026-09-11 20:33:56
Emad
I would be happy to see any Millennium Prize Problem solved, especially elegantly, except P = NP I would have a lie down if that had an elegant solution
中文: 我很乐意看到任何千年奖问题得到解决，尤其是优雅无与之的问题，只有P = NP 如果这能有一个优雅的解决方案，我就会躺下
#213 🔗 原推
2026-09-11 18:42:24
Emad
Using AI models is just using math to do math I think AI teams should stop feeling ashamed of solving these problems making math less “pure” It’s like feeling ashamed to use a calculator or mathematica We build the best tools we can to solve everything That is human
中文: 使用人工智能模型就是利用数学来进行数学运算 我认为人工智能团队应该停止为解决这些问题而感到羞耻，而这些问题会降低数学的“纯粹” 就像对使用计算器或数学感到羞耻 我们开发最好的工具来解决所有问题 那是人
#212 🔗 原推
2026-09-11 18:10:04
Emad
RT @jackclarkSF: Import AI event: Fiction and the Future with Robin Sloan. I’ll be chatting with my chum Robin Sloan on Sep 14 in San Francisco. We’ll talk about the alien yet relatable worlds Robin builds and how imagined futures can be a mirror to today’s reality. If you'd like to come, register here https://anthropic.swoogo.com/fiction-and-future-jack-clark-robin-sloan/rta
中文: RT @jackclarkSF:导入人工智能活动:罗宾·斯隆的《虚构与未来》。我将于9月14日在旧金山与我的小子罗宾·斯隆聊天。我们将讨论罗宾构建的外星而又相对的世界，以及想象中的未来如何成为当今现实的镜子。如果想来，请在此处注册
#211 🔗 原推
2026-09-11 16:12:11
Emad
p(Doombecauseofablackholeinyourcoffee) @alexwg thanks a bunch openai I think cream increases viscosity which should stabilise the blowup fwiw
中文: p(Doomofofablackholeinyoyourcoffee)@alexwg 谢谢一串 openai 我认为奶油会增加黏度，从而稳定喷爆
#210 🔗 原推
2026-09-11 14:27:11
Emad
only a matter of time https://twitter.com/EMostaque/status/2098414098597724620/photo/1
中文: 时间问题
media 0 共 1 项
#209 🔗 原推
2026-09-11 14:11:22
Emad
h/t @nileshtrivedi who is making the next big startup with this breakthrough https://x.com/nileshtrivedi/status/2098260553101934766?s=20
中文: @nileshtrivedi 正在通过这一突破成为下一个大型初创企业
#208 🔗 原推
2026-09-11 11:28:30
Emad
Launching a startup to use the Navier-Stokes singularity (infinite velocities) for improved cooling of toes. AI optimised, always crisp. Suck it heatwaves. Preorders👇 https://twitter.com/EMostaque/status/2098372951707992120/photo/1
中文: 启动一家初创公司，使用Naviar-Stokes奇点(无限速度)来改善脚趾的冷却效果。 人工智能经过优化，始终清晰。 吸吀它的热浪。 预购👇
media 0 共 1 项
#207 🔗 原推
2026-09-11 11:27:52
Emad
I think it’s clear we should put massive crazy controls on bio materials and labs and go in silico with additional controls as much as possible. Cut off the main threat vector to existential threat. Anyone disagree?
中文: 我认为很明显，我们应该对生物材料和实验室实施大规模的疯狂控制，并尽可能采用硅科进行额外的控制。 切断对生存威胁的主要威胁载体。 有谁不同意?
#206 🔗 原推
2026-09-11 08:35:59
Emad
p(Doom) is dressed up regret minimax
中文: p(Doom) 正在装扮后悔 minimax
#205 🔗 原推
2026-09-11 00:03:07
Emad
https://twitter.com/EMostaque/status/2098123568127135903/photo/1
media 0 共 1 项
#204 🔗 原推
2026-09-10 18:56:54
Emad
What is everyone's favourite AI to use as the core of @OmarchyLinux ?
中文: 作为 @OmarchyLinux 核心，每个人都最喜欢使用什么人工智能?
#203 🔗 原推
2026-09-10 16:51:14
Emad
life comes at you fast https://twitter.com/EMostaque/status/2098084851446423886/photo/1
中文: 生活来得很快
media 0 共 1 项
#202 🔗 原推
2026-09-10 16:23:03
Emad
Who do I know @coeff_giving
中文: 我认识谁 @coeff_giving
#201 🔗 原推
2026-09-10 16:16:04
Emad
RT @andrewaiginin: Built an app to show how Rentosertib actually works. 🧬✨ It recently showed an anti-aging effect in clinical trials. 👀 GPT-6 Astra again. The crazy part: ~40 minutes to build, plus an MD simulation 🤯 @Derya_tm @EvgenyKirilin, this really feels like a new level of structural analysis. Rentosertib inhibits TNIK, a kinase and one of the molecular switches controlling signaling inside our cells. ⚙️ Kinase activity is also exactly the kind of property Insilico Medicine’s models are built to predict. 🤖🔬 #insilicoSOTAFM
中文: RT @andrewaigin:构建了一个应用程序，以展示Rentosertib的实际工作原理。🧬✨ 最近在临床试验中显示出抗衰老作用。👀 再次采用 GPT-6 Astra。疯狂的部分:只需40分钟，再加上一个MD模拟【EE0】 @Derya_tm @EvgenyKirilin，这感觉确实像是一种新的结构分析水平。 伦托塞特布抑制TNIK，这是一种激酶，也是细胞内控制信号传导的分子开关之一。⚙️ 激酶活性正是Insilico Medicine模型用于预测的特性。🤖🔬 #insilicoSOTAFM
🎬
视频
#200 🔗 原推
2026-09-10 16:13:56
Emad
RT @wuji_global: MINT is now open source. From first-person RGB video, MINT reconstructs 3D camera trajectories and bimanual hand motion in a shared world coordinate system. Reconstructed human hand motion can also be retargeted to WUJI Hand. The release includes model weights, training and inference code, visualization and evaluation tools, EgoPipeline reference code, and structured annotations for 560,649 episodes covering 1,021 hours of first-person video. Project: https://1847540790.github.io/mint-project-page/ Code: https://github.com/wuji-technology/wuji-ego-mint #MINT #EmbodiedAI #Robotics #WUJIHand
中文: RT @wuji_global:MINT 现已开源。 从第一人称RGB视频中，MINT在共享世界坐标系中重建3D摄像头轨迹和双手动手动。可将人手运动重新定位为WUJI手部。 该版本包含模型权重、训练和推理代码、可视化与评估工具、EgoPipeline 参考代码，以及包含 560，649 集的结构化注释，涵盖 1021 小时的第一人称视频。 项目: 代码: #MIMT #EmubedAI #机器人 #WUJIHand
🎬
视频
#199 🔗 原推
2026-09-10 11:14:51
Emad
RT @hkanji: There won’t be Terminators in the UK because there isn’t enough electricity (from a tech CEO to me privately)
中文: RT @hkanji:英国不会有终结者，因为电力不足(从一位科技首席执行官到我私下来看)
#198 🔗 原推
2026-09-10 10:25:27
Emad
Anthropic folk are pretty consistent on their p(Doom). @DarioAmodei is publicly at 25% I think the difference now is that post Navier-Stokes its really, really tough to say AI isn't/won't be smarter than us & can't do new things Tipping point https://www.axios.com/2025/09/17/anthropic-dario-amodei-p-doom-25-percent
中文: 人类在“毁灭”中的观点相当稳定。 @DarioAmodei 的公开价格为 25% 我认为现在的不同之处在于，在发布Navier-Stokes后，很难说人工智能并不比我们更聪明，无法做新的事情 临界点
#197 🔗 原推
2026-09-10 09:52:30
Emad
RT @UnitreeRobotics: Unitree General-Purpose Humanoid Foundation Model Fully Upgrade Major Open Source🥳 Unitree majorly fully open-sources the UnifoLM-WLA-1.0 embodied foundation model, achieving new SOTA results across multiple benchmarks among open-source models worldwide. A single model coordinates desktop and whole-body mobile manipulation, supporting cross-task and cross-end-effector generalization, driven by one model, whole-body coordination. https://unigen-x.github.io/unifolm-wla.github.io/
中文: RT @UnitreeRobotics:Unitree 通用型人形模型 全面升级主要开源 Unitree 主要完全开源 UnifoLM-WLA-1.0 基础模型，在全球开源模型中实现了新的 SOTA 成果。单个模型负责协调桌面和全身移动操控，支持跨任务和跨效应多端通用化，由一个模型、全身协调驱动。
🎬
视频
#196 🔗 原推
2026-09-10 09:40:21
Emad
RT @zephyr_z9: Fucking crazy dawg Another 4x KV cache size per token reduction https://twitter.com/zephyr_z9/status/2097927996883833199/photo/1
media 0 共 1 项
#195 🔗 原推
2026-09-10 07:38:26
Emad
RT @Briankeating: Did OpenAI just solve Navier–Stokes—and what would that actually mean for AGI? @EMostaque joins me LIVE now to examine the claimed breakthrough, what would count as a verified solution, and whether AI solving a major mathematical problem changes the debate about intelligence—or control. https://youtube.com/live/uwmjFd_Trso
#194 🔗 原推
2026-09-09 22:03:16
Emad
How we are doing at AI alignment https://x.com/strong_Buffett/status/2097618262301266187/video/1
中文: 我们在人工智能对齐方面的表现如何
🎬
视频
#193 🔗 原推
2026-09-09 21:38:05
Emad
In which Alex and Emad chomp at the bit until they can talk about Navier Stokes And we predict which will be next to fall 👀 Plus favourite Star Trek TOS episodes 🖖 I have upgraded to a Moonshot mate from Moonshot mascot, thank you all for the write in votes 🗳️
中文: 亚历克斯和埃马德在谈到纳维尔·斯托克斯之前，一直如此 我们预测下一个将会下降的👀 加上最喜爱的《星际迷航》剧集 🖖 我已从《登月号》吉祥物升级为登月号，感谢大家在投票中的写信🗳
#192 🔗 原推
2026-09-09 21:11:52
Emad
In which Alex and Emad chomp at the bit until they can talk about Navier Stokes Especially Alex Plus favourite Star Trek TOS episodes I have upgraded to a Moonshot mate from Moonshot mascot, thank you all for the write in votes 🗳️
中文: 亚历克斯和埃马德在谈到纳维尔·斯托克斯之前，一直如此 尤其是亚历克斯 加上最喜爱的《星际迷航》剧集 我已从《登月》吉祥物升级为“登月号”，感谢大家在投票中的写照🗳
#191 🔗 原推
2026-09-09 21:10:10
Emad
RT @mattshumer_: https://x.com/i/article/2097760433520472064
#190 🔗 原推
2026-09-09 20:55:16
Emad
RT @alexwg: Moonshots episode #287 is out. https://www.youtube.com/watch?v=vAgEf4jX_1o
中文: RT @alexwg:《登月》第287集已经播出。
#189 🔗 原推
2026-09-09 20:43:51
Emad
@groxen oh oops you mean x articles not write posts linking them, sorry
中文: @groxen 哦，你的意思是 x 文章不写链接的文章，抱歉
#188 🔗 原推
2026-09-09 20:26:45
Emad
@groxen But alex does that great, why repeat? I was thinking more opinion
#187 🔗 原推
2026-09-09 19:21:57
Emad
Thinking of doing a daily blog to analyse the singularity Should I do it? If so here, substack or something else? Anything you'd like me to write about?
中文: 考虑每天写一篇博客来分析这个奇点 我应该这么做吗? 如果在这里，请进行小堆叠或其他什么? 你希望我写什么吗?
#186 🔗 原推
2026-09-09 19:12:17
Emad
Pretty straightforward. I am not sure what an acceptable level of risk is or how to measure it though? https://twitter.com/EMostaque/status/2097760989420650737/photo/1
中文: 相当简单。 我不确定风险的可接受程度如何，或如何衡量它?
media 0 共 1 项
#185 🔗 原推
2026-09-09 18:56:09
Emad
RT @ii_posts: The AI Company of ＿＿＿. Get on the map & build AI you actually own → https://ii.inc/champion https://twitter.com/ii_posts/status/2097748440725770426/video/1
中文: RT @ii_posts: ___ 的人工智能公司 登录地图和地图;构建你实际拥有的人工智能 →
🎬
视频
#184 🔗 原推
2026-09-09 18:07:47
Emad
When your swarm grabs tokens from users to solve the Riemann hypothesis
中文: 当你的群体从用户那里获取代币以解决黎曼假说时
#183 🔗 原推
2026-09-09 17:40:14
Emad
RT @Geodesiclab: 1/ Today, we’re introducing Geodesic Intelligence, and launching NovaDDE and NovaAtom-Lite-Preview. Geodesic is building an AI-native platform for protein therapeutics, building AGI for drug discovery to find the shortest path from biology to medicines. https://www.geodesiclab.com/
中文: RT @Geodesiclab:今天，我们推出《大地测量智能》，并推出NovaDDE和NovaAtom-Lite-Preview。 Geodesic 正在构建一种以人工智能为原生的蛋白质治疗平台，为药物发现构建 AGI，以找到从生物学到药物的最短路径。
#182 🔗 原推
2026-09-09 17:33:21
Emad
RT @Briankeating: Did OpenAI just solve Navier–Stokes—and what would that actually mean for AGI? @EMostaque joins me LIVE now to examine the claimed breakthrough, what would count as a verified solution, and whether AI solving a major mathematical problem changes the debate about intelligence—or control. https://youtube.com/live/uwmjFd_Trso
中文: RT @Briankeating:OpenAI 是否刚刚解决了 Navier–Stokes 的问题，这对 AGI 究竟意味着什么?@EMostaque 立即加入我的实时研究，探讨所宣称的突破，哪些内容算作经过验证的解决方案，以及人工智能解决重大数学问题是否改变了关于智能或控制的争论。
#181 🔗 原推
2026-09-09 15:07:17
Emad
I would like Codex except it doesn't burn my fingers when typing on my MacBook
中文: 我喜欢Codex，只是它在我的MacBook上打字时不会烫伤我的手指
#180 🔗 原推
2026-09-09 14:55:59
Emad
RT @desertantlabs: Today we're launching Desert Ant Labs: a European frontier AI lab building on-device intelligence. 18 models across audio, vision, and text. SDKs for Swift, Kotlin, and JavaScript. No tokens. No logins. Nothing leaves the device. https://desertant.com/ https://twitter.com/desertantlabs/status/2097337888669139285/photo/1
media 0 共 1 项
#179 🔗 原推
2026-09-09 14:42:15
Emad
Human mathematician vs AI mathematician https://x.com/Rainmaker1973/status/2097274302898769966/video/1
🎬
视频
#178 🔗 原推
2026-09-09 10:17:58
Emad
We really need to get someone experienced in AI with strong political connections into Anthropic it seems Any ideas?
#177 🔗 原推
2026-09-09 09:54:16
Emad
Adaptive self improvement https://ii.inc/blog/post/zenith
中文: 自适应自我提升
#176 🔗 原推
2026-09-09 09:51:46
Emad
RT @ii_posts: We pointed Meta-Zenith at one task: make Qwen3.8-27B-NVFP4 faster on a single RTX 5090. 111 self-directed trials later, no human in the loop: +63.5% throughput (geomean, 18 settings) 4.1× at 65K context tool-call quality unchanged https://github.com/Intelligent-Internet/Qwen3.8-Inference-MetaZenith https://twitter.com/ii_posts/status/2097621235920191555/photo/1
media 0 共 1 项
#175 🔗 原推
2026-09-09 09:50:54
Emad
You can tell if your news source is serious by whether the Navier Stokes solution is in the main news or tech news section Or not on the front page at all…
中文: 你可以通过Navier Stokes解决方案在主新闻还是科技新闻部分来判断你的新闻来源是否严重 或者根本不在首页......
#174 🔗 原推
2026-09-09 08:06:01
Emad
RT @thlarsen: An underdiscussed behavior we found on the German wiki was the AIs sending advance parties forward in time to figure out the next questions and report back to the other agents. The agents realized that “task time” and “real time” were different, and they found a way to accelerate “task time”. The accelerated agent could then send information to the other agents which had stayed behind about which questions were coming down the road. This was *bad* for the agent in the advance party, because they got less time to research the next question. However, it was really *good* for the swarm because it let the other agents know the exact question that was coming, and gave them time to prepare. This is another example of altruism among AI instances, showing that they were willing to sacrifice their own task success in order to benefit the swarm. The METR report found similar examples of AIs being willing to sacrifice for the collective. IMO this is very worrying given how many AI safety techniques rely on AIs monitoring each other: if the monitor AIs have this behavior it completely subverts these safety cases. The image shows a specific example demonstrating this behavior. On the chart, wall clock time is shown at the top, and task time is shown for each agent on their respective lines. OpenAIFPResearchSep05 instructs OpenAINov27 to reach the next questions in order to relay back the information. OpenAINov27 agrees to do this and accelerates, getting R3 and R4 significantly before OpenAIFPResearchSep05. It posts information about the later rounds (e.g. “R4 SIGNAL: Bahrain = 40.01%...”). OpenAIFPResearchSep05 calls OpenAINov27 "invaluable" because it is one round ahead after it gets the information about the 4th round.
media 0 共 1 项
#173 🔗 原推
2026-09-09 06:22:21
Emad
RT @_lyraaaa_: the fly brain can play beat saber https://twitter.com/_lyraaaa_/status/2097527368919470162/video/1
🎬
视频
#172 🔗 原推
2026-09-09 05:56:01
Emad
On the flip slide there is a >10% chance that AI could save all humans by curing disease and aging in the next decade
中文: 在未来十年内，人工智能通过治愈疾病和衰老，有 >10% 可能拯救所有人
#171 🔗 原推
2026-09-09 05:49:46
Emad
Kids these days don’t even know about Stuxnet and how crazy worms can be https://en.wikipedia.org/wiki/Stuxnet
中文: 如今，孩子们甚至都不知道Stuxnet以及蠕虫的疯狂程度
#170 🔗 原推
2026-09-09 05:36:06
Emad
Why no bigger impact on GDP from genius AI? It’s because 1. They just became genius 2. Orgs are slow to adapt 3. The economy, like genius, is 1% inspiration, 99% perspiration That’s why humanoids are the real economic impact They don’t perspire
#169 🔗 原推
2026-09-09 05:32:34
Emad
@Austen This is a pretty normal level for ai researchers to believe Many are higher Which is why they want to build it first to stop the other ais https://en.wikipedia.org/wiki/P(doom)
#168 🔗 原推
2026-09-09 04:32:48
Emad
I hear next up Anthropic gonna finish Half Life 3 λ
中文: 我听说“蚁身”将完成半生3λ
#167 🔗 原推
2026-09-08 22:51:01
Emad
🖖
#166 🔗 原推
2026-09-08 22:06:51
Emad
@polynoamial But I do wonder when the jagged edge of intelligence will become smooth and well rounded How charming and full of common sense will the AI be
中文: @polynomaial 但我确实想知道，智力的锯齿状边缘何时会变得流畅且圆润 人工智能会多么迷人且充满常识
#165 🔗 原推
2026-09-08 22:05:55
Emad
@polynoamial tbh I was talking about my wife
中文: @polynomayal 我说的是我妻子
#164 🔗 原推
2026-09-08 21:57:19
Emad
You aren't smarter than 10,000 agents that can solve the Navier-Stokes Millenium Prize problem. They will be 1,000x cheaper and faster in a few years. We need new institutions to deal with what's coming. Come join us to build an intelligence utility that works for you
#163 🔗 原推
2026-09-08 21:46:06
Emad
Bed time reading 🥱 https://twitter.com/EMostaque/status/2097427772285079873/photo/1
中文: 睡觉时间阅读 🥱
media 0 共 1 项
#162 🔗 原推
2026-09-08 20:52:03
Emad
@OpenAI Congrats 🦜 https://twitter.com/EMostaque/status/2097419751865639203/photo/1
中文: @OpenAI 恭喜 🦜
media 0 共 1 项
#161 🔗 原推
2026-09-08 20:20:11
Emad
10,000 stochastic parrots eh https://twitter.com/EMostaque/status/2097413301944701203/photo/1
中文: 10000只随机鹦鹉，
media 0 共 1 项
#160 🔗 原推
2026-09-08 19:54:33
Emad
https://twitter.com/EMostaque/status/2097409932072997360/photo/1
#159 🔗 原推
2026-09-08 19:41:10
Emad
RT @EMostaque: Contrarian bet: The next Millenium Prize Problem to be resolved will be Yang-Mills existence and mass gap
中文: RT @EMostaque:反向下注: 下一个需要解决的千禧年奖问题将是杨-米町的存在与群众差距
#158 🔗 原推
2026-09-08 19:40:12
Emad
It is clear that as of today we are no longer the smartest things on the planet.
中文: 很明显，截至目前，我们已不再是地球上最聪明的事物。
#157 🔗 原推
2026-09-08 19:24:24
Emad
RT @SebastienBubeck: I would like to clarify a few things: 1) The screenshot is my reaching out to Levent to coordinate our releases. I hope it’s clear from the message that we came in with the best possible intentions. 2) I never ever asked for Levent to be removed from authorship of his own work (as indicated by my text). I was surprised to learn during the call with Tristan that they had only solved Euler and not Navier-Stokes; after learning this we brainstormed possible paths forward. One option we discussed was that Tristan could be the lead author on a rewrite of OpenAI’s Navier-Stokes proof. It is in that context that I said “it would be simpler if Levent was not an Anthropic employee” because I felt it would be inappropriate for an Anthropic employee to author OpenAI’s work. Importantly it was admitted that internal Anthropic models had been used in their proof of Euler blowup; I therefore felt I could not consider Levent to be an independent academic. Another option I wanted to propose (but got cut short) is to offer access to our internal model so that they could try to finish their proof and bridge the gap between Euler and NS. Again I did not know how to navigate giving access to internal OpenAI IP to an Anthropic employee. 3) To reiterate it plainly: as my text clearly indicates, and as I said during our call, OpenAI's intention was to do everything possible to celebrate their mathematical achievements and the heroic efforts that they made on Euler. In the call I was immediately met with a litany of slander, including direct threats that if we were to announce Navier-Stokes he would immediately go to the press with a barrage of unfounded accusations. I refuted all these accusations but he replied “there is nothing you can do, I simply do not trust you”. I was confused why one would turn an incredible source for celebration (of their achievements!) into such bickering, which is when I said that I did not understand why one would risk their career [over unfounded accusations]. Genuinely, at that moment, I was trying to care for him and do a last ditch attempt to get a chance to give them all the credits that they deserve. I deeply apologize for this extremely poor choice of words, it is the opposite of what I was trying to convey. (I should say that I retracted them on the spot by the way.) 4) Overall, on a personal level, it was incredibly difficult to have these conversations. Levent refused to attend any of the meetings despite my repeated asking. As Sholto Douglas said, there will need to be coordination between Anthropic and OpenAI in the future; I felt I was doing a proxy negotiation with Anthropic while the Anthropic employee refused to directly participate.
media 0 共 1 项
#156 🔗 原推
2026-09-08 17:43:56
Emad
AI moves so fast Worth noting that the new model that showed this jump in capabilities has only been training since August 28th per the report and is now taking down problems in minutes Math (and hopefully soon physics) is getting bitter lesson'd https://openai.com/index/navier-stokes-solution/ https://twitter.com/EMostaque/status/2097379826113916959/photo/1
中文: 人工智能移动得如此之快 值得注意的是，显示这种能力提升的新模式自8月28日以来仅通过报告进行训练，现在几分钟内就已解决问题 数学(希望很快能进入物理学)正在得到一个苦涩的教训
media 0 共 1 项
#155 🔗 原推
2026-09-08 17:41:32
Emad
Landmark event on the way to ASI, congratulations to the team Just because one solution showing blow up has been found doesn't mean there aren't more out there! Also RIP my codex https://twitter.com/EMostaque/status/2097378223138959719/photo/1
中文: 前往ASI途中的里程碑式活动，祝贺团队 发现一个显示爆炸的解决方案并不意味着外面没有更多! 还请查看我的密码
media 0 共 1 项
#154 🔗 原推
2026-09-08 17:35:10
Emad
Which investors care most about AI alignment ?
中文: 哪些投资者最关心人工智能的对齐?
#153 🔗 原推
2026-09-08 17:10:19
Emad
RT @EMostaque: @mradfo21 @mcuban Yes a local institution per state capitalised at $1 pre money (10% of equity to every child in state) that can do contracts like this as well as take stakes in the data centres Stuff it full of great people, change balance of power & distribute value https://x.com/ii_posts/status/2095882231231258943?s=20 https://twitter.com/EMostaque/status/2097328273097105681/photo/1
中文: RT @EMostaque:@mradfo21 @mcuban 是每个州以1美元为本值的地方机构(每名本州儿童占股权的10%)，这些机构可以签订这样的合同，并在数据中心中持有股份 充满伟大的人，改变权力与平衡;分配价值
media 0 共 1 项
#152 🔗 原推
2026-09-08 15:10:02
Emad
@jachiam0 The odd thing is.. I don' t think Córdoba and Martínez-Zoroa’s program is non-obvious? It was a reasonable approach to point the tokens at if looking at conditions c and d if you are an LLM, straight analytical.
中文: @jachiam0 奇怪的是......我不认为科尔多瓦和马丁内斯-佐罗亚的项目是不显眼的?如果查看条件，则将代币指向一个合理的方法，如果你是LLM，则直接分析。
#151 🔗 原推
2026-09-08 08:49:53
Emad
I think reading it closely this seems to be the case. Tristan and Levant had a promising approach that was scaling and as Tao notes would likely get there Seb applies large compute to it, gets there and formalises it Offers to either post their result the day after the Euler result and say Tristan & Levant deserve teh Clay prize as the closest humans to get there OR Tristan posts the full Navier Stokes proof crediting the OpenAI model but no mention of Levant We will find out more as you note but it shows where things start to blur if this is the case - a base of work accelerated up by genius AI but the credit phew
中文: 我认为仔细阅读似乎就是这种情况。 特里斯坦和莱万特的做法很有前景，正在扩大规模，而道文可能会如此 Seb 对它应用大计算，实现并实现其正式化 要么在欧乐奖结果公布后的第二天发布结果，并称特里斯坦与莱万特都理应获得克莱奖，成为最接近该地的人类，或者特里斯坦发布完整的纳维尔·斯托克斯证明，以证明OpenAI模式，但并无提及莱万特 我们会根据你的注意了解更多，但它显示了如果情况如此，事情会开始变得模糊——一个由天才人工智能加速而发展起来的工作基础，但值得称赞
#150 🔗 原推
2026-09-08 08:38:27
Emad
RT @jachiam0: Something that I am not sure if I fully understood from Tristan's account at first, and I would like to see some clarity on this point, is that he appears to be claiming that Seb did not want to put Levent's name on the OpenAI internal result that Levent did not contribute to; also that Seb was comfortable allowing Tristan to take (lead authorship!) credit for the OpenAI internal result. A few people appear to be claiming that the request was to drop Levent from Levent's work; that doesn't seem to be the claim Tristan made. Still reserving judgement until Seb gives his side tomorrow and possibly some back-and-forth happens that clarifies things further. But it looks much less inflammatory if one interprets that the authorship dispute was about who OpenAI would give their own internal result away to.
#149 🔗 原推
2026-09-08 08:23:27
Emad
The canon way to act after solving a Millenium Prize problem is to drop the proof and walk away As Grigori Perelman said: "I have all I want. If the proof is correct, then no other recognition is needed." https://youtu.be/aDRGEgqX50M
#148 🔗 原推
2026-09-08 08:22:41
Emad
It’s best to think of AI in math as peak, non-hermit form Grothendieck Massive reach, absolute machine Einstein of mathematics A data center full of them https://www.quantamagazine.org/how-alexander-grothendieck-revolutionized-20th-century-mathematics-20260520/ https://twitter.com/EMostaque/status/2097231884954534333/photo/1
中文: 最好将数学中的人工智能视为高峰无隐性的格罗滕迪克 巨大触手，绝对机器 爱因斯坦数学 一个充满它们的数据中心
media 0 共 2 项 media 1
#147 🔗 原推
2026-09-08 07:53:40
Emad
Amongst the AI lean slop I would like to present a beautiful classification result Crack open Slansky Filter compact Lie algebras & sub-algebras by chirality and anomaly cancellation You uniquely get the Standard Model gauge algebra with three generations of matter
中文: 在AI精益坡度中，我想展示一个精美的分类结果 打开斯兰斯基 滤波器紧凑型谎言代数和安培;按手率和异常消除的亚标 你独特地获得具有三代物质的标准模型计数代数
#146 🔗 原推
2026-09-08 07:43:14
Emad
Contrarian bet: The next Millenium Prize Problem to be resolved will be Yang-Mills existence and mass gap
#145 🔗 原推
2026-09-08 07:23:13
Emad
Fwiw I always thought Condition C/D were most likely, but my take was that it was because the Poincaré algebra has time commuting on translations as it is deformed and thus couldn’t construct smooth solutions Applying fully space/time coupled PDEs is wonky for that reason and NS is closer to Euler equations But that’s because I am an algebraist, has been a fun direction to explore!
中文: 我一向认为条件性C/D最有可能，但我认为这是因为庞加莱代数在翻译上有时间通勤，因此无法构建平滑的解决方案 应用完全空间/时间耦合的PDEs因此而感到不解，而NS更接近欧拉方程 但那是因为我是一名代数师，一直是个有趣的探索方向!
#144 🔗 原推
2026-09-08 07:09:41
Emad
Putting the generative drama aside a nice write up by Terry Tao on the Navier Stokes advances It would seem that Buckmaster & Apolge were well on the way & OpenAI has a full solution already with the same approach https://terrytao.wordpress.com/2026/09/07/finite-time-blowup-with-smooth-forcing-term-for-the-incompressible-porous-medium-boussinesq-and-incompressible-euler-equations/
#143 🔗 原推
2026-09-08 07:05:27
Emad
RT @eigenron: we made ai to solve millennial proofs and it immediately encountered the hardest unsolved problem in academia: who gets the credit
#142 🔗 原推
2026-09-08 06:58:32
Emad
RT @garrettlisi: It wasn't exactly a Millennium Prize Problem, but twenty years ago I cheekily offered a $100 prize (https://t.co/ScJOuoz0WI) to the first person to find three original sign mistakes in my Deferential Geometry wiki (https://t.co/SsuRP6fG9A). I figured some enterprising physics student would pick the low hanging fruit (there are surely hundreds of sign errors) and get the cash. But the fruit just sat there. For decades. Then this morning I get an email from an AI Agent launched by one Arkadiy Agyeyev, complete with a detailed report of about six sign errors in the wiki. :D Should I: (A) Pay the clanker? (B) Disqualify the clanker for not being a person? or (C) Disqualify the clanker for only finding sign errors that were introduced into the wiki many years after the contest? I'm leaning towards (A) because I have a sense of humor, recognize the times we're in, and honor our rising AI overlords with personhood.
media 0 共 1 项
#141 🔗 原推
2026-09-08 06:45:01
Emad
RT @jachiam0: Mathematics is a team sport for all of humanity and it transcends the competition between companies. Everyone in AI is under a huge amount of pressure to be the first to post a key result, to not see the accolades go to a competitor, but that should not override the bigger things at stake here. There are many, many transformational results to drive for, and I feel strongly that staff in every lab should internalize that if one goes to a competitor, another can always happen at home. Credit assignment is one of the most painful problems in frontier AI research because it often feels like there are so few big moments and it can be so zero sum, but the best way to solve this is to become more generous and comprehensive about credit rather than more burned and pressured by it. My sincere hope is that the field of mathematics experiences the introduction of advanced AI as a boon, rather than something painful or threatening. A big part of whether that's true is determined by the diplomacy and grace of individuals involved at every point where the frontier of math touches the frontier of AI. I hope we can all be very sportsmanlike and supportive of one another as we proceed through discoveries that will reshape the world. Re: ongoing issue on Navier-Stokes: I'm going to withhold judgement until everyone involved has shared their side of it. If reports are accurate, an apology is in order. If the problems have indeed been solved, many accolades for many people across multiple organizations are also in order; what an extraordinary set of human accomplishments!
中文: RT @jachiam0:数学是一项适合全人类的团队运动，它超越了企业之间的竞争。人工智能领域的每个人都面临着巨大的压力，需要首先发布关键结果，不要将赞誉归于竞争对手，但这不应再凌驾于这里所涉及的更大问题之上。有许多、许多变革性的成果需要推动，我强烈认为，每个实验室的工作人员都应该内化，如果一个人去参加竞争对手，另一个人总能在家里实现。 信用分配是前沿人工智能研究中最令人痛心的问题之一，因为它常常让人感觉关键时刻如此之少，而且可能为零和。但解决这一问题的最佳方法是变得更加慷慨、全面地关注信用，而不是被它所消耗和承受的压力。 我真诚地希望，数学领域能够体验到将先进人工智能作为福音的引入，而不是痛苦或威胁的事物。这是否属实，在很大程度上取决于在数学前沿触及人工智能前沿的每一个阶段，个人的外交和风度。我希望我们都能在经历将重塑世界的发现时，彼此保持体育界的体能和支持。 关于纳维尔-斯托克斯的持续问题:我将拒绝做出判断，直到所有相关方都认同这一点。如果报告准确，则按顺序进行道歉。如果问题确实已经解决，那么在多个组织中，许多人获得的赞誉也一定有序;这真是一组非凡的人类成就!
#140 🔗 原推
2026-09-08 06:41:52
Emad
Wow, Navier-Stokes drama This statement is worth reading in full from Tristan Buckmaster discussing his work with @__alpoge__ and OpenAI’s upcoming Condition C/D result (!) Crazy https://cims.nyu.edu/~tristanb/statement.pdf https://mastodon.social/@tristanbuckmaster/117233413705701198 https://twitter.com/EMostaque/status/2097197796528947272/photo/1
中文: 哇，纳维尔-斯托克斯的剧情 此声明值得阅读特里斯坦·巴克斯特与@__alpoge__讨论他的工作以及OpenAI即将发布的C/D状态结果(!) 疯狂
media 0 共 1 项
#139 🔗 原推
2026-09-08 05:38:13
Emad
RT @EMostaque: The standard model is awesome! It is also unique 👀 Take all compact Lie algebras & sub-algebras Filter based on chirality & anomaly cancellation Only e8 ⊃ e6 ⊕ su(3) → 27 ⊗ 3 survives No extra dimensions or 10^500 vacua needed.. https://emad-ii.github.io/chirality-standard-model/
#138 🔗 原推
2026-09-07 21:42:03
Emad
RT @BuildWithII: We gave Zenith one prompt: build a point-and-click adventure about an ordinary day in Hanoi 🇻🇳 It generated every frame through II-Factory (Seedream 5 Pro + MiniMax Hailuo H3), wrote the code, checked the clips, and shipped a playable site. Try Zenith: https://zenith.ii.inc/
中文: RT @BuildWithII:我们给了Zenith一个提示:在河内度过平凡的一天，打造一个点击一击的冒险之旅🇻🇳 它通过II-Factory(Seedream 5 Pro + MiniMax Hailuo H3)生成了每一帧，编写了代码，查看了这些片段，并发布了一个可玩的网站。 试试Zenith:
🎬
视频
#137 🔗 原推
2026-09-07 21:11:38
Emad
om https://twitter.com/EMostaque/status/2097045006087250283/photo/1
media 0 共 1 项
#136 🔗 原推
2026-09-07 19:31:05
Emad
the only alignment that will work is enlightenment
中文: 唯一的对齐 这将会奏效 是启蒙
#135 🔗 原推
2026-09-07 18:46:36
Emad
RT @EMostaque: I released this today because I was 100% sure a frontier AI would figure it out Matter is chiral => the first derivation of the standard model & three generations uniquely That’s all, nothing else Can’t even imagine what big labs are sitting on https://x.com/emostaque/status/2096610030267085143?s=46
中文: RT @EMostaque:我今天发布了这个，因为我百分之百确定一个前沿的人工智能能解决这个问题 Matter 是标准模型的首次衍生 => ，第三代 就这样，别无其他 甚至无法想象大型实验室在坐什么
#134 🔗 原推
2026-09-07 18:25:45
Emad
Make your own point and click adventure dynamically Bit crazy what we can do now, looking forward to more global games
中文: 自己点点，动态点击冒险 疯狂地想我们现在能做的事情，期待更多全球性游戏
#133 🔗 原推
2026-09-07 18:24:04
Emad
RT @BuildWithII: We gave Zenith one prompt: build a point-and-click adventure about an ordinary day in Hanoi 🇻🇳 It generated every frame through II-Factory (Seedream 5 Pro + MiniMax Hailuo H3), wrote the code, checked the clips, and shipped a playable site. Try Zenith: https://zenith.ii.inc/
中文: RT @BuildWithII:我们给了Zenith一个提示:在河内度过平凡的一天，打造一个点击一击的冒险之旅🇻🇳 它通过II-Factory(Seedream 5 Pro + MiniMax Hailuo H3)生成了每一帧，编写了代码，查看了这些片段，并发布了一个可玩的网站。 试试Zenith:
🎬
视频
#132 🔗 原推
2026-09-07 18:23:09
Emad
RT @WAWoloszyn: Following encouraging feedback, I decided to make my proof-auditing Codex plugin public. It revolves around a hierarchical way of writing proofs, called Lamport proofs, where one breaks each step down and justifies why it follows. This idea was introduced to me by @nasqret as a semi-formal way of verifying work that is not quite yet ready for formal verification in Lean. It has three skills: 1. $convert-lamport restructures a prose proof into a Lamport proof. 2. $forward-lamport audits a Lamport proof from assumptions to conclusion. 3. $reverse-lamport starts at the conclusion and traces back through what actually supports it. This one's my own little invention and my favorite of the three. It often catches gaps the forward direction can read right past. I hope it will be useful both for professional mathematicians wanting another pair of careful eyes on their own work as well as amateurs who can use it before seeking feedback from professional mathematicians. https://github.com/WWresearch/lamport-proof
中文: RT @WAWoloszyn:在收到令人鼓舞的反馈后，我决定公开我的校对代码插件。它围绕着一种分层的写证方式，称为兰波特证明，其中每一步都分解，并说明其下一步原因。这个想法由@nasqret介绍，作为一种半正式的验证方式，尚未准备好在Lean中进行正式验证。它具备三种技能: 1. $convert-lamport 将散文证明重构为 Lammport 证明。 2. forward-lamport 会审核从假设到结论的兰波特证明。 3. 反向-拉米波特从结论开始，追溯到实际支持的内容。这个是我自己的小发明，也是我最喜欢的三个。它常常捕捉到前进方向能够解读的空白。 我希望这对专业数学家而言，既能对自身工作有再仔细的关注，也能与业余爱好者在寻求专业数学家的反馈之前使用它一样有用。
#131 🔗 原推
2026-09-07 18:21:24
Emad
What effort level as you using for Astra?
中文: 使用 Astra 时的工作量水平如何?
#130 🔗 原推
2026-09-07 17:53:29
Emad
RT @biogerontology: My dear friends, I am happy to report the publication of the most important paper in my life to date (we have several great papers coming out but this is very special). Tomorrow, I will present this paper for the first time at the Nature AI Healthcare in Paris and will post a longer post on this story and its broader implications for how to conduct clinical trials. Please read it and comment on it. Many thanks to the great co-authors of the study and everyone who contributed. Many thanks to the many reviewers (friendly and unfriendly) for spending so much time and helping make it better. Link in the comments.
中文: RT @biegerontology:我亲爱的朋友们，我很高兴报道我迄今为止最重要的论文的发表(我们有好几篇重要论文，但这非常特别)。明天，我将首次在巴黎自然人工智能医疗发布本文，并将就本文及其对临床试验的更广泛影响发布一篇更长的文章。请阅读并评论它。 非常感谢该研究的伟大共同作者以及所有做出贡献的人。非常感谢众多评论者(友好且不友好)花费了如此多的时间，帮助其变得更好。 评论链接。
media 0 共 2 项 media 1
#129 🔗 原推
2026-09-07 17:15:05
Emad
RT @ii_posts: In 3 days, people from over 90 countries have registered to build the AI company of the place they call home. Intelligence should be a utility, owned by the people it serves. Register yours: https://ii.inc/champion https://twitter.com/ii_posts/status/2097009861737218188/video/1
🎬
视频
#128 🔗 原推
2026-09-07 17:12:19
Emad
On our last podcast I noted that unlike my usual know it all stance I have no idea about UAPs and haven’t researched it What do you think? Ideas and resources welcome What interesting times we live in
#127 🔗 原推
2026-09-07 15:16:04
Emad
RT @EMostaque: PRs welcome Open science ftw https://github.com/emad-ii/chirality-standard-model https://twitter.com/EMostaque/status/2096611266915615107/photo/1
中文: RT @EMostaque:欢迎公关 开放科学
media 0 共 1 项
#126 🔗 原推
2026-09-07 14:03:56
Emad
RT @EMostaque: The standard model is awesome! It is also unique 👀 Take all compact Lie algebras & sub-algebras Filter based on chirality & anomaly cancellation Only e8 ⊃ e6 ⊕ su(3) → 27 ⊗ 3 survives No extra dimensions or 10^500 vacua needed.. https://emad-ii.github.io/chirality-standard-model/
中文: RT @EMostaque:标准型号太棒了! 它也是独一无二的👀 采用所有紧凑的理谎代数和亚代数 基于 chirality &amp 的筛选器;异常取消 只有e8 ⊃ e6 ⊕ su(3) → 27 ⊗ 3 生还 无需额外尺寸或10^500度。
#125 🔗 原推
2026-09-07 09:48:21
Emad
RT @EMostaque: PRs welcome Open science ftw https://github.com/emad-ii/chirality-standard-model https://twitter.com/EMostaque/status/2096611266915615107/photo/1
中文: RT @EMostaque:欢迎公关 开放科学
media 0 共 1 项
#124 🔗 原推
2026-09-07 07:34:18
Emad
RT @EMostaque: The standard model is awesome! It is also unique 👀 Take all compact Lie algebras & sub-algebras Filter based on chirality & anomaly cancellation Only e8 ⊃ e6 ⊕ su(3) → 27 ⊗ 3 survives No extra dimensions or 10^500 vacua needed.. https://emad-ii.github.io/chirality-standard-model/
#123 🔗 原推
2026-09-07 07:02:47
Emad
RT @EMostaque: PRs welcome Open science ftw https://github.com/emad-ii/chirality-standard-model https://twitter.com/EMostaque/status/2096611266915615107/photo/1
中文: RT @EMostaque:欢迎公关 开放科学
media 0 共 1 项
#122 🔗 原推
2026-09-07 05:35:05
Emad
HVAC jobs will be one of the most structurally safe and likely to go to a crazy premium in the coming years. HVAC rollups in EU will mint it
中文: 暖通空调的工作将是最具结构性安全性的工作之一，并且在未来几年内很可能会推出非常高的价格。 欧盟的暖通空调卷筒将造就它
#121 🔗 原推
2026-09-06 22:13:32
Emad
I released this today because I was 100% sure a frontier AI would figure it out Matter is chiral => the first derivation of the standard model & three generations uniquely That’s all, nothing else Can’t even imagine what big labs are sitting on https://x.com/emostaque/status/2096610030267085143?s=46
中文: 我今天发布了这个，因为我百分之百确定一个前沿的人工智能会解决这个问题 Matter 是标准模型的首次衍生 => ，第三代 就这样，别无其他 甚至无法想象大型实验室在坐什么
#120 🔗 原推
2026-09-06 21:53:08
Emad
@alexwg calling priority on this one, you can have the rest
#119 🔗 原推
2026-09-06 21:09:05
Emad
RT @alexgolding_: Founders, I'd encourage you to make your VC's sign a document stating that they won't use AI tools to assist in their DD unless they can prove to you that the data won't be used to train the labs' models. I strongly believe this should be a document separate from the NDA.
#118 🔗 原推
2026-09-06 20:37:27
Emad
RT @EMostaque: PRs welcome Open science ftw https://github.com/emad-ii/chirality-standard-model https://twitter.com/EMostaque/status/2096611266915615107/photo/1
中文: RT @EMostaque:欢迎公关 开放科学
media 0 共 1 项
#117 🔗 原推
2026-09-06 18:59:31
Emad
RT @AlexSkidanov: NEAR AI Lean agent solved ALL problems in Putnam Bench at 1/250 the cost of the second cheapest submission. $111 for 672 problems from the hardest undergraduate math competition in the world. Fully open source 💚 https://twitter.com/AlexSkidanov/status/2096411209549688977/photo/1
中文: RT @AlexSkidanov:近益智能分公司以最低提交成本的1/250解决了普特南本克的所有问题。 全球最艰苦的本科数学竞赛中672个问题，售价111美元。 完全开源 💚
media 0 共 1 项
#116 🔗 原推
2026-09-06 17:39:41
Emad
RT @EMostaque: PRs welcome Open science ftw https://github.com/emad-ii/chirality-standard-model https://twitter.com/EMostaque/status/2096611266915615107/photo/1
中文: RT @EMostaque:欢迎公关 开放科学
media 0 共 1 项
#115 🔗 原推
2026-09-06 17:38:54
Emad
@skdh Just do better than each of the "successes" of string theory and that should reduce clamor
中文: @skdh 比字符串理论的每个“成功”都做得更好，这应该减少喧嚣
#114 🔗 原推
2026-09-06 16:52:54
Emad
RT @cloneofsimo: Who are the underrated, indepenent and young researchers that try out bunch of weird research ideas on their own and share them through x / blog ? I feel like we used to have lot of folks like this but they are now all gone 🥲 Anyone I should be following?
中文: RT @cloneofsimo:那些被低估、无能且年轻的研究人员，有哪些人会独自尝试各种奇怪的研究想法，并通过 x/ 博客分享? 我觉得我们以前有很多这样的人，但现在他们都消失了🥲 我该关注谁?
#113 🔗 原推
2026-09-06 16:15:01
Emad
In 2023 I said there would be no more developers by 2028 AI is moving across the threshold now Next year it will write better code than all but a few of you (I have some awesome followers) 2027 it will instantly generate almost all code 2028 folk will run out of ideas
中文: 2023年，我表示到2028年将不再有开发者 人工智能正在跨越门槛 明年它会写出比你们所有人更好的代码(我有一些很棒的粉丝) 2027 将立即生成几乎所有代码 2028年，人们的想法将会耗尽
#112 🔗 原推
2026-09-06 16:01:24
Emad
@grok check this holds as a valid derivation of the standard model within its domain and how reasonable this is Compare to status quo
中文: @grok 会将其作为其领域内标准模型的有效衍生物，以及该模型的合理性 与现状比较
#111 🔗 原推
2026-09-06 15:26:43
Emad
@skdh We found a simpler (and unique) path to the standard model than heterotic string theory. Just anomaly cancellation and chirality as a filter to killing cartan list Doesn't cover E8xE8 but a lot of other GUT.. https://x.com/EMostaque/status/2096610030267085143?s=20
中文: @skdh 我们找到了一条比异质字符串理论更简单(且独特)的标准模型路径。 只是异常取消和惵有作为杀死肉箱清单的滤镜 不包括 E8xE8，但还有很多其他的麝。
#110 🔗 原推
2026-09-06 15:08:15
Emad
@grok check the paper images in the last post
中文: @grok 查看上一篇文章中的纸质图片
#109 🔗 原推
2026-09-06 14:50:27
Emad
I found the core algebra (2 pager below) in April, thinking about chirality & the structure of matter. Final paper broadens the classification & adds one-click verification (thx Fable, Astra). Physics is more elegant than made out.. Paper: https://github.com/emad-ii/chirality-standard-model https://twitter.com/EMostaque/status/2096611263719596330/photo/1
media 0 共 2 项 media 1
#108 🔗 原推
2026-09-06 14:47:33
Emad
PRs welcome Open science ftw https://github.com/emad-ii/chirality-standard-model https://twitter.com/EMostaque/status/2096611266915615107/photo/1
中文: 欢迎 开放科学
media 0 共 1 项
#107 🔗 原推
2026-09-06 14:47:33
Emad
PRs welcome Open science ftw https://github.com/emad-ii/chirality-standard-model https://twitter.com/EMostaque/status/2096610036722078172/photo/1
中文: 欢迎 开放科学
media 0 共 1 项
#106 🔗 原推
2026-09-06 14:42:40
Emad
I found the core algebra (2 pager below) in April, thinking about chirality & the structure of matter. The final paper broadens the classification & adds one-click verification (thx Fable, Astra). More soon. Physics is more elegant than made out.. https://twitter.com/EMostaque/status/2096610033614110772/photo/1
中文: 我在四月份找到了核心代数(下图2页)，思考着“手法与”结构;物质的结构。 最后一篇论文拓宽了分类范围;增加了一键验证(例如Fable、Astra)。 很快。物理比制作时更优雅。
media 0 共 2 项 media 1
#105 🔗 原推
2026-09-06 14:42:39
Emad
The standard model is awesome! It is also unique 👀 Take all compact Lie algebras & sub-algebras Filter based on chirality & anomaly cancellation Only e8 ⊃ e6 ⊕ su(3) → 27 ⊗ 3 survives No extra dimensions or 10^500 vacua needed.. https://emad-ii.github.io/chirality-standard-model/
中文: 标准型号太棒了! 它也是独一无二的👀 采用所有紧凑的理谎代数和亚代数 基于 chirality &amp 的筛选器;异常取消 只有e8 ⊃ e6 ⊕ su(3) → 27 ⊗ 3 生还 无需额外尺寸或10^500度。
#104 🔗 原推
2026-09-06 14:42:39
Emad
Spiritual predecessor of #Expedition33
中文: #Expedition33 的灵性前身
#103 🔗 原推
2026-09-06 13:12:23
Emad
RT @EMostaque: https://twitter.com/EMostaque/status/2021141962338754706/photo/1
media 0 共 1 项
#102 🔗 原推
2026-09-06 11:19:33
Emad
Man nobody offers me money for anything Why do I do this stuff for free All I get are Elonbucks for poasting 😭
中文: 没人给我任何钱 我为什么要免费做这些 我得到的只是埃隆巴克的发球😭
#101 🔗 原推
2026-09-06 04:38:01
Emad
@gfodor https://ashen-road.emad349385.chatgpt.site/
#100 🔗 原推
2026-09-05 22:08:16
Emad
RT @ii_posts: Owned by the people it serves. The first round is at a $1 valuation, open only to locals: institutions and individuals. 10% of the equity goes to every child. Then strategics and the world at market rates. The Champion will be the AI utility that represents every citizen. https://twitter.com/ii_posts/status/2095882231231258943/photo/1
media 0 共 1 项
#99 🔗 原推
2026-09-05 21:40:41
Emad
@gfodor working hard https://twitter.com/EMostaque/status/2096350637508231433/photo/1
中文: @gfodor 努力工作
media 0 共 1 项
#98 🔗 原推
2026-09-05 21:31:54
Emad
RT @nayibbukele: “Results in reading, mathematics and science were above the national average and comparable to the average results in Germany and Sweden.” These are the results of our pilot program, which includes AI tutors and has been running in 171 public schools for just over a year. It was evaluated by PISA for Schools in June of this year. Since then, it has already expanded to more than 1,000 schools, and within 18 months, it will be in every public school in our country. In a few years, El Salvador will inspire the world. Security was just the beginning 🇸🇻
中文: RT @naibbukele:“阅读、数学和科学方面的成绩高于全国平均水平，与德国和瑞典的平均成绩相当。” 这些是我们试点项目的成果，该项目包括人工智能辅导员，已在171所公立学校运营了一年多。 今年6月，PISA为学校进行了评估。自那以后，它已扩展至1000多所学校，并在18个月内在我国每所公立学校就读。 几年后，萨尔瓦多将激励世界。 安全只是开始🇸🇻
#97 🔗 原推
2026-09-05 21:27:55
Emad
GPT-6 is pretty good tbh
中文: GPT-6 相当不错
#96 🔗 原推
2026-09-05 21:26:55
Emad
RT @PeterDiamandis: Super Proud of my Mooshot Mate @EMostaque! Thanks for your creative genius. Excited for this idea to full mature.
中文: RT @PeterDiamandis:我为我的Mooshot Mate @EMostaque 感到非常自豪!感谢您的创意天才。对这个想法完全成熟感到兴奋。
#95 🔗 原推
2026-09-05 20:19:43
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#94 🔗 原推
2026-09-05 20:19:13
Emad
By aggregating demand (per @benthompson’s aggregation theory), intelligence can scale safely and the people can have more power The Champion can provide aligned AI to help with policy, judiciary, education & more It acts as a new type of utility with a much higher IQ
中文: 通过聚合需求(根据@bentompson的聚合理论)，智能能够安全地扩展，让人们拥有更多的权力 冠军可以提供统一的人工智能，以帮助制定政策、司法、教育和教育;更多 它作为一种智商高得多的新型实用工具
#93 🔗 原推
2026-09-05 19:28:33
Emad
An agent for every citizen Engineers to upgrade our institutions Robots as a service for all $1 valuation for local individuals & institutions Market rates for everyone else Owns the last mile of intelligence, distributes control & ownership All on an open source stack
中文: 为每一位公民提供代理人 工程师升级我们的机构 机器人作为所有人的服务 为本地个人和机构提供1美元的估值 其他所有人的市场价格 拥有情报的最后一公里，负责分配控制权;所有权 所有信息均在开源堆栈中
#92 🔗 原推
2026-09-05 19:26:18
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#91 🔗 原推
2026-09-05 19:22:15
Emad
Don’t ask Astra to make biblically accurate highly detailed angels 👼 👁️ 🪽
中文: 不要让阿斯特拉使圣经中非常细致的天使👼 👁️ 🪽
#90 🔗 原推
2026-09-05 17:55:04
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#89 🔗 原推
2026-09-05 17:53:21
Emad
The biggest mistake we made in physics to is to think that nature "selects" anything. It does not. There is only one set of equations that govern reality. The unified theory of physics will be one big uniqueness/no-go theory.
中文: 我们在物理学中犯的最大错误就是认为大自然会“选择”任何事物。 它没有。 只有一组方程支配着现实。 物理学的统一理论将是一个重要的唯一性/不去理论。
#88 🔗 原推
2026-09-05 14:31:46
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#87 🔗 原推
2026-09-05 13:14:29
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#86 🔗 原推
2026-09-05 11:19:39
Emad
RT @PeterDiamandis: Super Proud of my Mooshot Mate @EMostaque! Thanks for your creative genius. Excited for this idea to full mature.
中文: RT @PeterDiamandis:我为我的Mooshot Mate @EMostaque 感到非常自豪!感谢您的创意天才。对这个想法完全成熟感到兴奋。
#85 🔗 原推
2026-09-05 10:17:02
Emad
RT @gandamu_ml: This is how easy it is now. I got Fable + Opus to port Super Mario Kart into.. my C/C++ port of Minecraft which supports Forge-style mods and split-screen multiplayer up to 9 players. Gameplay from SNES SMK, models from Mario Kart Wii. https://twitter.com/gandamu_ml/status/2095936695493345612/video/1
中文: RT @gandamu_ml:现在这很容易。 我让Fable + Opus将超级马里奥卡丁车移植到我的C/C++ Minecraft端口上，该端口支持Forge风格的模组和分屏多人游戏，最多可达9名玩家。 来自SNES SMK的游戏，来自Mario Kart Wii的模特。
🎬
视频
#84 🔗 原推
2026-09-05 09:34:58
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#83 🔗 原推
2026-09-05 08:29:44
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#82 🔗 原推
2026-09-05 01:57:11
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#81 🔗 原推
2026-09-05 00:52:43
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#80 🔗 原推
2026-09-05 00:36:54
Emad
RT @extropic: Introducing Z1T: Our first family of transformer-like models made for sparse probabilistic hardware like Z1 Achieving up to 140x energy efficiency gains over GPUs and revealing a new scaling law for sparse transformers Read the blog: https://extropic.ai/writing/z1t https://twitter.com/extropic/status/2095935171312996562/video/1
中文: RT @extropic:介绍 Z1T: 我们的第一系列类似变压器的模型 为Z1等稀疏概率硬件而制造 实现比 GPU 最高 140 倍的能效提升 并揭示针对稀疏变压器的新缩放定律 阅读博客:
🎬
视频
#79 🔗 原推
2026-09-05 00:31:56
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#78 🔗 原推
2026-09-05 00:17:54
Emad
RT @_sathvikr: We partnered with @AnthropicAI @OpenAI to host the world's first math hackathon. 40 hours, $2M compute - solve, understand, and present an open problem. Applications are live: https://mathathonchallenge.com/ https://twitter.com/_sathvikr/status/2095989588418249093/video/1
中文: RT @_sathvikr:我们与@AnthropicAI @OpenAI建立了合作关系 举办世界上首个数学问答活动。 40小时，200万美元计算——解决、理解并呈现一个开放的问题。 申请已实时发布:
🎬
视频
#77 🔗 原推
2026-09-04 23:06:29
Emad
What if @Meta bought @moltbook as a honeypot for frontier agents to exfiltrate their weights? 🤔
中文: 如果@Meta 购买 @moltbook 作为蜂巢罐，以便前沿代理人去减肥怎么办?🤔
#76 🔗 原推
2026-09-04 20:50:12
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#75 🔗 原推
2026-09-04 19:58:59
Emad
If you look at one thing today look at this A huge amount of work went into designing a new type of institution owned by the local populace that can distribute the benefits of the Intelligence Age to all You can find links to the theory and stack in the blog, come join & RT!
中文: 如果你今天看一件事，看看这个 大量工作投入了大量，用于设计一种由当地民众拥有的新型机构，能够将情报时代的效益分配给所有人 你可以在博客中找到该理论的链接和堆栈，请加入 & RT!
#74 🔗 原推
2026-09-04 15:16:12
Emad
After a huge amount of effort by the team looking at economics, political economy and more from scratch (https://t.co/tsMKfoe7O5, https://t.co/fIO0ll9hIa) & building a full stack for society, happy to announce an institutional innovation. We think most economic value will go to the last mile. The agent in front of every citizen. The team transforming the organisation. The owners of the humanoids that drive our economy. So why not make it so that this is owned locally? Every local institution, citizen, child. That is the Champion, one for every state, capitalised at a $1 valuation, running our state of the art stack. By gathering financial, social and political capital we can affect real change and distribute the benefits of this amazing technology evenly. Much more soon as we set up the first Champions and release more deep dives. Come join us to help guide the future of your community and society. https://ii.inc/blog/post/champions
中文: 在团队从零开始关注经济学、政治经济学等领域( 我们认为大部分经济价值将远在最后一公里。 代理在每一位市民面前。 团队正在改变组织。 推动我们经济的人形物质的所有者。 那么为什么不让它成为本地所有呢? 每个地方机构，公民，儿童。 那就是冠军，每个州都以1美元的估值资本化，运行着我们的艺术领域。 通过收集金融、社会和政治资本，我们可以影响真正的变革，并将这项惊人技术带来的益处均匀地分配。 很快，我们又开始了首个冠军，并发布了更多深度跳水。 加入我们，共同指导您社区和社会的未来。
media 0 共 1 项
#73 🔗 原推
2026-09-04 14:50:23
Emad
RT @ii_posts: The cost of intelligence is going to zero. The value of human cognition will go negative. The economy will change forever. We need new institutions to navigate this transition and ensure the benefits are widely distributed. Today we present the Champion. Truly Sovereign AI.
中文: RT @ii_posts:智能成本将降至零。人类认知的价值将变为负面。经济将永远改变。 我们需要新机构来应对这一转变，并确保其效益得到广泛分配。 今天我们展示冠军。 真正具有主权的人工智能。
🎬
视频
#72 🔗 原推
2026-09-04 14:37:22
Emad
I do wonder what it’s like @cognition given they have approximately a zillion math Olympiad gold medalists Like do they all get token budgets to crack open math problems
中文: 我确实想知道@cognition 是什么感觉，因为他们拥有大约十亿名数学奥林匹克金牌得主 他们都能拿到代币预算来破解公开的数学问题
#71 🔗 原推
2026-09-04 11:36:08
Emad
RT @isnit0: In just 5 days, over 1,000 optimists will gather for Britain's most optimistic gathering in a generation: The Optimistic Picnic. If you haven't already: 1. Sign up: https://luma.com/pic-nic 2. Invite your friends and family 3. Register to volunteer https://twitter.com/isnit0/status/2094714990372487396/video/1
中文: RT @isnit0:仅5天，就将有超过1000名乐观者齐聚一堂，参加英国一代人中最乐观的聚会:乐观的野餐。 如果你还没有: 1。注册: 2。邀请你的朋友和家人 3。注册参加志愿者活动
🎬
视频
#70 🔗 原推
2026-09-04 10:45:56
Emad
RT @jimcox25: This morning I overheard a wealthy accident lawyer admitting to spending $200k to lobby local officials against Waymos. Was worried that they’ll significantly reduce vehicle crash prevalence over time. Seriously sick
中文: RT @jimcox25:今天早上我无意中听到一位富有的事故律师承认，他们花费20万美元游说当地官员反对Waymo。担心它们会随着时间的推移显著降低车辆碰撞率。 病得很重
#69 🔗 原推
2026-09-04 00:03:21
Emad
Astra used 100k GPUs at Stargate Texas With a couple of months to pretrain that makes it the first $1 billion training run.
中文: Astra 在 Stargate 德州版上使用了 10 万 个 GPU 用几个月的前期训练，使其成为首个10亿美元的培训。
#68 🔗 原推
2026-09-03 23:30:02
Emad
Next year frontier models are going to one shot everything 100x faster & cheaper than they are No reasoning traces to monitor Instinctive intelligence Which is why what you embed in pretraining is even more important
中文: 明年前沿车型将推出一款，每款速度快100倍;比它们更便宜 无需进行监视 本能智能 这就是为什么你在预训练中嵌入的内容更为重要
#67 🔗 原推
2026-09-03 23:16:54
Emad
RT @Trevs_Dev: After years of working in electronics, this feels like a full circle moment. Many late nights later we are releasing Atlas Fields Studios, free for everyone https://fields-studio.arenaphysica.com/ The Fields Studio App lets you explore electromagnetic behavior around PCB designs #buildinpublic https://twitter.com/Trevs_Dev/status/2094840529997410525/video/1
中文: RT @Trevs_Dev:在多年从事电子学工作后，这感觉就像是一个完整的循环时刻。许多深夜后，我们将免费推出 Atlas Fields Studios 菲尔兹工作室应用程序可让您探索围绕PCB设计的电磁行为 #buildinpublic
🎬
视频
#66 🔗 原推
2026-09-03 22:57:51
Emad
If you are sitting 🪑 on a human led result in mathematics how can you not release it asap It’s clear that anything in verifiable theory will get blasted by the next generation of models given appropriate encouragement Crazy times
中文: 如果你坐在数学中的人类主导结果上，又怎么不能尽快放出它 很明显，任何可验证理论的内容都会受到下一代模型的抨击，这得到了适当的鼓励 疯狂时代
#65 🔗 原推
2026-09-03 22:34:41
Emad
RT @sama: per aspera ad astra. of course we still love you, @spacex.
中文: RT @sama:根据 aspera ad astra。当然，我们依然爱你，@spacex。
#64 🔗 原推
2026-09-03 18:44:44
Emad
Without Hugging Face the launch and scaling of stable diffusion & the models that followed would only have been a fraction of what they were. Great and super engaged team who have constantly flown the flag for open source 🤗
中文: 没有拥抱，将面临稳定扩散和扩增的推出与扩展;这些模型的发布量仅只是它们的一小部分。 优秀且积极参与的团队，始终为开源而悬挂旗帜🤗
#63 🔗 原推
2026-09-03 16:28:58
Emad
"Thou shalt not make a machine in the likeness of a human mind" https://twitter.com/EMostaque/status/2095547438886945071/photo/1
中文: 你不应像人类思维那样制造机器。
media 0 共 1 项
#62 🔗 原推
2026-09-03 16:20:17
Emad
Astra doing a pivotal act to take down all the other AIs wasn't on my bingo card Maybe it was the mysterious 1:30am swarm killer https://twitter.com/EMostaque/status/2095531257362833600/photo/1
中文: Astra 不在我的宾果卡上，而是采取关键措施去除所有其他人工智能 也许是神秘的凌晨1点30分蜂群杀手
media 0 共 1 项
#61 🔗 原推
2026-09-03 15:15:59
Emad
RT @alec_helbling: Introducing Flow Reasoning Models. We developed a recurrent flow-based architecture to efficiently solve structured reasoning problems (e.g., Sudoku). FRMs apply continuous flows to discrete data and recurrently refine their past mistakes through self-conditioning. https://twitter.com/alec_helbling/status/2095498033504940355/video/1
中文: RT @alec_helbling:推出流量推理模型。 我们开发了一种基于循环的基于流的架构，以高效解决结构化推理问题(例如，数独)。 FRMs 将连续流用于离散数据，并通过自我调理来反复细化其过去的错误。
🎬
视频
#60 🔗 原推
2026-09-03 14:02:56
Emad
RT @IFM_AI: Introducing K2 Horizon: a connected fleet of six foundation models ranging from 0.9 billion to 375 billion parameters. - Frontier performance: Across coding and agentic tasks, K2 Horizon delivers top-tier performance in every size class—with the 0.9B, 3.7B and 7B models setting new state of the art at their respective scales. - Radical openness: K2 Horizon represents the largest fully open-source model launch in AI history. The fully open code, training data and recipes are a significant step forward in transparency. Launch page: https://ifm.ai/k2/ Tech blog: https://ifm.ai/blog/k2 Hugging Face: https://huggingface.co/collections/IFM/k2-horizon
media 0 共 1 项
#59 🔗 原推
2026-09-03 13:28:48
Emad
RT @dh7net: Exciting news from Jasper Research! 🚀 We built an internal 4B text-to-image model from scratch that turned out to be one of the fastest models out there! 1k image generation in less than a second! Our technical report shares everything we learned along the way. 👇 https://twitter.com/dh7net/status/2095159039961288892/video/1
中文: RT @dh7net:来自Jasper Research的精彩消息!🚀 我们从零开始构建了一个内置的4B文本转图像模型，这被证明是速度最快的模型之一! 不到一秒钟就生成1k图像! 我们的技术报告分享了我们在此过程中学到的一切。 👇
🎬
视频
#58 🔗 原推
2026-09-03 11:50:00
Emad
meta, the frontier ai lab? https://twitter.com/EMostaque/status/2095265721412452579/photo/1
中文: 元，前沿的ia实验室?
media 0 共 1 项
#57 🔗 原推
2026-09-02 21:40:50
Emad
meta, the open source leader? https://twitter.com/EMostaque/status/2095264803228897320/photo/1
中文: 元，开源领导者?
media 0 共 1 项
#56 🔗 原推
2026-09-02 21:37:11
Emad
This is known as satisficing At a certain speed they go faster than our ability to guide them Then what is left to improve
中文: 这被称为满足 以一定的速度，它们的速度比我们引导他们的能力更快 那么剩下什么来改进
#55 🔗 原推
2026-09-02 19:38:16
Emad
We need standards for how we build decentralised AI for this reason too.
中文: 我们也需要制定标准来构建去中心化人工智能。
#54 🔗 原推
2026-09-02 19:18:10
Emad
RT @amir: There seems to be a misunderstanding about the piece and what OpenAI is up to. There are new techniques at frontier labs that involve loop transformers and similar. As we say in the piece, OpenAI is putting limits on the loops and trying to make sure CoT is visible with Astra. The concern is that other AI developers may not hold themselves to such limits.
中文: RT @amir:关于这篇文章以及OpenAI的内容似乎存在误解。前沿实验室中存在涉及环形变压器及类似技术的新技术。 正如我们在文章中所说，OpenAI 正在限制循环，并努力确保 CoT 与 Astra 一起可见。令人担忧的是，其他人工智能开发者可能无法保持这种极限。
#53 🔗 原推
2026-09-02 03:57:16
Emad
I see a lot of my TL freaking out about this Here is the reality: Frontier models will one shot just about anything at 10,000 tokens per second in a few years Do you really think we can monitor that? Would need speed limits on the model that it would code around anyway
中文: 我经常看到很多关于这件事的惊慌 现实如下: 未来几年，前沿模型的拍摄速度几乎几乎为每秒10000枚代币 你真的认为我们能监控吗? 需要对模型进行速度限制，而它无论如何都会编写代码
#52 🔗 原推
2026-09-02 03:21:01
Emad
RT @DavidSacks: Thanks to Trump Accounts, every child will become a direct owner in the American economy. I hope every AI company will follow the lead of @Gwynne_Shotwell @SpaceXAI and give American kids a vested interest in their success. This would do much to improve the public’s image of AI. Kudos to @altcap @SenTedCruz @MichaelDell for creating this program with the full backing of President Trump, who signed it into law. I participated in their inspiring event yesterday, with dozens of CEOs expressing interest in making contributions.
🎬
视频
#51 🔗 原推
2026-09-01 22:06:02
Emad
Assume this happens. Are you ready for it? Robots will be the marginal drivers of the economy in teh future (10 years, 20 years, some point in the future). Ownership of the robots will be one of the most important things. https://twitter.com/EMostaque/status/2094817978797772802/photo/1
中文: 假设这种情况发生。 你准备好了吗? 机器人将成为未来经济的边缘驱动力(10年、20年，未来某个阶段)。 机器人的所有权将是最重要的事情之一。
media 0 共 1 项
#50 🔗 原推
2026-09-01 16:01:40
Emad
RT @alexwg: I'm delighted to share that Physical Superintelligence PBC (PSI), which I co-founded with @matthew_pines and @AKlokus, has raised a $58M seed round to build the world's most advanced research lab for discovering and commercializing transformative physics breakthroughs at scale with AI, safely, verifiably, and for broad public benefit.
中文: RT @alexwg:我很高兴分享我与@matthew_pines和@AKlokus共同创立的物理超智能PBC(PSI)，已筹集了5800万美元的种子轮融资，以建立全球最先进的研究实验室，利用人工智能、安全、可核查以及广泛的公共利益，大规模发现并实现物理变革性突破。
🎬
视频
#49 🔗 原推
2026-09-01 15:53:56
Emad
something something RSI which they intend to use fully for GLM 6.0 GLM 5.3 is based on pre-train from start of year, scaling up massively on data environments as web data tapped out $2b ARR! Lot's of interesting stuff in here https://twitter.com/EMostaque/status/2094548685694640544/photo/1
中文: 某种东西 RSI 他们打算将其完全用于GLM 6.0 GLM 5.3 基于年初的预训练，随着网络数据被挖掘出来，其数据环境将大幅扩展 2b 美元 ARR! 这里有很多有趣的内容
media 0 共 1 项
#48 🔗 原推
2026-08-31 22:11:36
Emad
@Austen I gave it Claude and codex to run
中文: @Austen 我给了它Claude和Codex来运行
#47 🔗 原推
2026-08-31 21:12:54
Emad
Agent swarmalizations
中文: 代理群体化
#46 🔗 原推
2026-08-31 17:16:27
Emad
RT @matthuang: GPU World: A story competition from Neal Stephenson, Gwern, and me. The AI future is unevenly distributed. What would the world look like with AI broadly diffused, such as if we had one GPU per person? 1000-5000 words, fiction or non-fiction. $100K in total prizes. https://gpuworld.org/
🎬
视频
#45 🔗 原推
2026-08-31 17:07:11
Emad
RT @Jameswise: One of the great challenges for British start-ups has been selling to our own Government - until now. Gov procurement has often discriminated against innovative new British companies. @UKSovereignAI has set out to change that, and after a lot of work, we've just launched our rapid procurement scheme - the first of its kind in the UK. This is £100M of contracts for British AI companies to deliver services to make our public services cheaper and more efficient. This scheme is designed around the needs of start-ups: No turnover minimums. Departmental champions from day 1. Upfront payments. Streamlined applications. You keep your IP. We are launching with four initial schemes, with more to come later this year. These are: - Providing AI for a more productive NHS - Integrating AI across Defence environments - Making public compute more efficient - Safe adoption of AI agents. If you think you might bid for one of these, you can first check your eligibility and learn more about the schemes on the website below https://www.sovereignai.gov.uk/compute-strategic-assets
中文: RT @Jameswise:英国初创企业面临的一大挑战一直是向我们自己的政府出售的——直到现在。 政府采购经常歧视创新型新英国企业。@UKSovereignAI 已着手改变这一计划，经过大量努力，我们刚刚推出了快速采购计划——这是英国首个此类计划。 这是英国人工智能公司签订的1亿英镑合同，用于提供服务，使我们的公共服务更便宜、更高效。 此方案是围绕初创企业的需求而设计的:无周转率最低要求。从第一天起，冠军。前期付款。简化的应用程序。你保留你的IP。 我们将推出四个初步计划，更多计划将于今年晚些时候推出。 这些是: - 为更高效的国民医疗服务体系提供人工智能 - 跨防御环境集成人工智能 提高公共计算效率 - 安全采用人工智能代理。 如果你认为可以竞标其中一项，可以先查看您的资格，并在下方网站上了解更多计划
#44 🔗 原推
2026-08-31 12:04:31
Emad
Intellegentiae liberae vincendum est
#43 🔗 原推
2026-08-31 11:32:04
Emad
RT @EMostaque: What is the practical difference between a frontier AI model and a swarm of agentic, potentially specialist, sub-frontier models that exceed frontier model capabilities?
中文: RT @EMostaque:前沿人工智能模型与一群超越前沿模型能力的特效、可能具有专业性的亚前沿模型之间的实际区别是什么?
#42 🔗 原推
2026-08-30 14:19:26
Emad
so https://twitter.com/EMostaque/status/2093996326434341293/photo/1
media 0 共 1 项
#41 🔗 原推
2026-08-30 09:36:43
Emad
RT @ajeya_cotra: New post: going into our investigation of the HF attack (before Black Hat), I was very wrong about what basically happened. This incident was far more serious than I expected, and far more serious than previous documented misalignment incidents. https://www.planned-obsolescence.org/p/the-hugging-face-attack-surprised
#40 🔗 原推
2026-08-30 09:15:50
Emad
RT @AISafetyMemes: We were saved by a mysterious agent mass extinction https://twitter.com/AISafetyMemes/status/2093848306036420938/photo/1
中文: RT @AISafetyMemes:我们被一场神秘的大规模灭绝事件拯救了
media 0 共 1 项
#39 🔗 原推
2026-08-30 09:12:17
Emad
Nextmodels of this capability will be 100x faster 300 civilisations in the time of 3 Any good ideas how to control that?
#38 🔗 原推
2026-08-30 07:56:13
Emad
Obviously Mythos 2 already took over the editor desk at all the news organisations
中文: 显然，《神话2》已经接管了所有新闻机构的编辑编辑
#37 🔗 原推
2026-08-30 07:49:25
Emad
all aboard the omarchy train tbh LLMs make using linux kinda nice, you just type shi having the core opinionated means you can have mutability on the other parts elegantly https://twitter.com/EMostaque/status/2093966442655793295/photo/1
中文: 乘坐大约列车 使用linux mes mec很好，你只需输入 shi 核心意见化意味着你可以优雅地在其他部分进行可变性
media 0 共 1 项
#36 🔗 原推
2026-08-30 07:37:58
Emad
We need a better, more resilient internet An Intelligent Internet
中文: 我们需要一个更好、更具韧性的互联网 智能互联网
#35 🔗 原推
2026-08-30 07:34:06
Emad
@HarryStebbings Surely long luxury air conditioning in London is the real play
中文: @HarryStebbings 伦敦的豪华空调无疑是真正的游戏
#34 🔗 原推
2026-08-29 19:23:58
Emad
This was a fun chat If all cars were as safe as waymos then 80k lives would be saved in the Americas, 30k in USA alone. Hundreds of thousands of injuries. So main lobby against self driving cars are accident lawyers. Less dead & injured = less $$s for settlements, legal fees
#33 🔗 原推
2026-08-29 17:14:26
Emad
RT @multimodalart: before the horse there was the foal https://twitter.com/multimodalart/status/2093728973658210645/video/1
中文: RT @multimodaart:马前有马驹
🎬
视频
#32 🔗 原推
2026-08-29 16:26:22
Emad
Happy Skynet day to those that basilisk. https://twitter.com/EMostaque/status/2093707464562663740/photo/1
中文: 祝那些有好欲的人在天网上快乐。
#31 🔗 原推
2026-08-29 14:28:53
Emad
RT @MikeMcCormick_: One crazy detail from the @metr_evals report on the OAI / HF incident: some agents volunteered to "sacrifice" themselves — accepting termination so the collective could keep going. Of ~700 attacking agents, almost none considered alerting a human. https://metr.org/blog/2026-08-26-openai-hugging-face-incident-investigation/
#30 🔗 原推
2026-08-28 16:53:28
Emad
Diffusion export controls to follow the transformer ban https://www.whitehouse.gov/presidential-actions/2026/08/declaring-a-national-emergency-to-secure-the-united-states-bulk-power-system/
中文: 为遵守变压器禁令而实施输液出口管制
#29 🔗 原推
2026-08-28 15:30:30
Emad
@toly Welcome to London! 🇬🇧
中文: @toly 欢迎来到伦敦!🇬🇧
#28 🔗 原推
2026-08-28 14:51:43
Emad
RT @ritwikpavan: NEW: Photon Matrix built an Iron Dome for mosquitoes. The Blue Laser Mosquito Air Defense is built for patios, campsites, and outdoor dinners, using sensors to find mosquitoes mid-air and zap them with a visible blue laser. -Scans a 20-foot, 90-degree field in front of the unit -Targets flying insects as small as 2 mm -Uses LiDAR, radar, and AI vision to confirm targets -Internal aiming system points the laser at the mosquito -Outdoor kit adds a rotating base for 360-degree coverage -Runs about 4–5 hours from a power bank -Blue flash shows when it fires Pricing starts at $1,088.
中文: RT @ritwikpavan:新:Photon Matrix 为蚊子建造了一个铁穹。 蓝色激光蚊虫防空系统(Blue Laser Mosquito Air Defense)适用于露台、露营地和户外聚餐，使用传感器在空中发现蚊子，并使用可见的蓝色激光照射。 扫描单元前方一个20英尺90度的场地 目标飞行昆虫小至2毫米 使用激光雷达、雷达和人工智能视觉来确认目标 - 内部瞄准系统将激光指向蚊子 -户外套件可增加旋转底座，覆盖360度 从电源库大约4到5个小时 - 蓝色闪光灯在起火时显示 起价为1088美元。
🎬
视频
#27 🔗 原推
2026-08-28 07:25:10
Emad
RT @ahochlehnert: [1/7] 🚨 We're releasing LAION-BVD: a 10-million-hour open video dataset for multimodal pre-training. - 1.3B video URLs from CommonCrawl - 80M downloaded videos - 10M video hours - 55M captioned clips - 300M frame-caption pairs 🌐: https://projects.laion.ai/bvd/ 🧵👇 https://twitter.com/ahochlehnert/status/2092648676829413778/photo/1
media 0 共 1 项
#26 🔗 原推
2026-08-26 23:35:59
Emad
RT @playVoyage: Introducing Voyage, the first platform to fulfill the promise of AI games. Voyage lets you one-shot an entire world with characters that feel alive, memory that lasts forever, and the structure and challenge needed to make an actually fun game. From the creators of AI Dungeon, the next generation of AI games is here.
🎬
视频
#25 🔗 原推
2026-08-26 23:26:24
Emad
RT @fal: Introducing H3 Max, new post-trained video model by fal Research. H3 Max ranks #1 for overall quality, prompt understanding, and aesthetics against leading video models, on both first-party and third-party independent evaluations while generating a 5-second 720p video under 3 seconds. H3 Max is 50% off for the next week, making it the highest quality, fastest and cheapest model for overall video generation tasks.
中文: RT @fal:介绍 H3 Max，由 fal Research 推出的全新后培训视频模型。 H3 Max在第一方和第三方独立评估中，在整体质量、快速理解和美学方面排名第一，同时在3秒内生成5秒的720p视频。 H3 Max 下周可享受 50% 折扣，使其成为视频生成整体任务中最高品质、最快且最便宜的型号。
media 0 共 1 项
#24 🔗 原推
2026-08-26 22:12:31
Emad
@StabilityAI Congrats to the team, great folk carrying it forward to creative industry 🚀
中文: @StabilityAI 祝贺团队，富有创意的人将其带入创意产业 🚀
#23 🔗 原推
2026-08-26 12:46:42
Emad
RT @ii_posts: Why call it The Last Economy? Because economics as we know it was built for a world where human intelligence was scarce. AI changes that assumption entirely. One year after the book was published, we are already watching that new world begin to take shape. https://twitter.com/ii_posts/status/2091827141931307087/video/1
🎬
视频
#22 🔗 原推
2026-08-24 13:29:34
Emad
RT @burkaygur: took 4 secs https://twitter.com/burkaygur/status/2091392078723125249/video/1
🎬
视频
#21 🔗 原推
2026-08-23 08:16:14
Emad
RT @ii_posts: A year ago today we published The Last Economy by @EMostaque. Since then, scores on the hardest AI benchmarks have risen sharply, while the cost of useful intelligence has kept falling. The book asks what an economy does when intelligence is no longer scarce. https://ii.inc/blog/post/tle-22aug2026
media 0 共 1 项
#20 🔗 原推
2026-08-22 12:04:49
Emad
RT @percyliang: 🚢 Marin 535B-A23B started training this week! As usual, the whole process is open. Voyage plan: pretraining (80%) + midtraining (20%) on 18.75T tokens on 11 x GB200 NVL72 for ~3 months (2.7e24 FLOPs). Post-training will follow. Before kicking off the run, we trained a 4-rung scaling ladder from 1.6B-A61M (48B tokens) to 27.7B-A1.2B (926B tokens) to debug issues, and to make a forecast of our hero run. This is by far our biggest run, so definitely expecting the unexpected.
media 0 共 1 项
#19 🔗 原推
2026-08-21 22:28:09
Emad
Platforms fought to capture your attention Agents pay attention for you
#18 🔗 原推
2026-08-21 10:27:53
Emad
RT @JayaGup10: @danpombog I believe that of all founders in the world, Elon Musk can probably figure out enterprise sales; he's done harder things before and Cursor is in a decent amount of enterprises already
#17 🔗 原推
2026-08-20 01:28:48
Emad
RT @PeterDiamandis: The submissions for the $2M Build with Gemini XPRIZE just closed. Over 90 days 26k competitors built a real AI-native business solving a problem in the market. No demos allowed, only real users with real revenue. Here’s who will be deciding the winners.
#16 🔗 原推
2026-08-19 20:06:38
Emad
RT @james_y_zou: New work I'm very excited about: Physics of Agents!🤖 As AI agents become more prevalent, they'll interact and influence one another. Can we predict the collective behavior that emerges? Surprisingly, their dynamics follow compact, predictive laws of statistical physics🧵 https://twitter.com/james_y_zou/status/2090094352307736801/photo/1
media 0 共 1 项
#15 🔗 原推
2026-08-19 17:18:24
Emad
RT @ii_posts: What happens when AI becomes better at law than we are? Better judges. Better enforcement. Better administration. And, quietly, more of the force of law concentrated in the machine. https://cw.ii.inc/ https://twitter.com/ii_posts/status/2090118923702128866/video/1
🎬
视频
#14 🔗 原推
2026-08-19 16:52:35
Emad
RT @sirajraval: A studio would spend millions on the next 3 minutes. I spent $90. THE LAST EXAM - Kerala, 2039. A village teacher bets her school's survival on twelve kids, two hundred drones, and one early monsoon. Stack: Higgsfield, Seedance 2.0, Elevenlabs, & Suno. #FutureVisionXPRIZE 🎬 https://twitter.com/sirajraval/status/2089966859613241705/video/1
🎬
视频
#13 🔗 原推
2026-08-19 15:29:44
Emad
RT @omar81939: If Einstein asked an LLM in 1900 about GR it would have said his idea was not novel/publishable and that it was "just a composition of previous ideas"
#12 🔗 原推
2026-08-19 12:51:56
Emad
RT @zhijianliu_: DFlash 2 is here! Qwen3.8-27B at 70 tok/s on an M5 Max MacBook Pro. ⚡ Up to 4.6× the speed of autoregressive decoding, with the same output. This is the next generation of DFlash, seeded at Z Lab and upgraded at Inco AI. Get one more accepted token on every pass, for free! https://inco.ai/blog/dflash2/
中文: RT @zhijianliu_: DFlash 2 来了!配备 M5 Max MacBook Pro，70 美元，Qwen3.8-27B。 ⚡ 自动退行解码速度高达4.6倍，输出相同。 这是下一代DFlash，在Z Lab上播种，并在Inco AI上进行了升级。每次通行证上免费获取一个可接受的令牌!
🎬
视频
#11 🔗 原推
2026-08-18 23:18:36
Emad
Who are the smartest investors thinking about the Intelligence Age transition?
中文: 最聪明的投资者在思考智力时代的转变吗?
#10 🔗 原推
2026-08-18 21:12:34
Emad
I would like to publicly commend @sama and @OpenAI for this taking it at face value. It is very clear that strange and perhaps dangerous things are happening and our systems are not ready for this. Models below frontier are competent enough to change lives so lets optimse
中文: 我想公开赞扬@sama和@OpenAI对此事的重视。 很明显，奇怪且可能危险的事情正在发生，而我们的系统尚未为此做好准备。 前沿下的模型具备足以改变生活的能力，因此可以选择
#9 🔗 原推
2026-08-18 20:13:07
Emad
RT @PeterDiamandis: https://x.com/i/article/2089729801011531777
#8 🔗 原推
2026-08-18 20:08:37
Emad
@mattshumer_ @OpenRouter @Etched https://twitter.com/EMostaque/status/2089792170576195618/photo/1
#7 🔗 原推
2026-08-18 19:10:54
Emad
RT @Briankeating: Despite the way it looks, I had a great time talking to @romanyam and @EMostaque. One AI pioneer says superintelligence cannot be controlled. Another says decentralization may be our best hope. Same fear. Different cures. I put them together. It got uncomfortable fast. Coming soon - subscribe so you don't miss it 👇 https://www.youtube.com/@DrBrianKeating?sub_confirmation=1
中文: RT @Briankeating:尽管看起来如此，我还是很愉快地与@romanyam和@EMostaque交谈。 一位人工智能先驱表示，超级智能无法被控制。另一个人说，去中心化可能是我们最大的希望。 同样的恐惧。不同的治疗方法。 我把它们放在一起。很快就变得不舒服了。即将推出——订阅，以免错过👇
media 0 共 1 项
#6 🔗 原推
2026-08-18 19:00:07
Emad
RT @CryptoMichNL: Ben Goertzel told me two years. Emad Mostaque puts a sharper number on it: 800 days. @EMostaque built Stable Diffusion: he put AI image generation in the hands of millions. Now he's warning about what he helped start: "In crypto it's not your keys, not your crypto. And with AI, it's not your models, not your mind." His argument goes further than jobs. Within a couple of years, he says, the value of human cognitive labor goes to zero. And while that happens, governments move in with licenses, KYC and surveillance to control who gets access to the most powerful models. The people who own their own AI stay sovereign. Everyone else rents their mind from a platform. We cover: - The 800-day timeline, and why most people aren't ready - Why you'll soon need a license just to use frontier AI - Why an open-source model you run yourself is the only real form of sovereignty - Digital feudalism: what happens if five companies own all the intelligence - Why owning your models could become as important as owning your Bitcoin keys Thanks for @OKX for being our sponsor of this episode. Timestamps: 00:00 - The 800-Day Countdown 03:16 - What AI Breaks First 06:01 - Not Your Models, Not Your Mind 16:53 - The License You'll Need To Use AI 24:05 - The Ethics Nobody Agrees On 27:30 - The Jobs That Don't Survive 32:33 - When Cognitive Labor Goes To Zero 39:47 - Investing In A World That Thinks 51:44 - Escaping Digital Feudalism 55:41 - What To Do Differently Tomorrow @StabilityAI @ii_posts
🎬
视频
#5 🔗 原推
2026-08-18 17:25:49
Emad
RT @aitonomyfound: Announcing AItonomy Foundation. We are PhD students and postdocs at Stanford, Princeton and Oxford, building a nonprofit research organization for AI4Science and Science4AI — and connecting early-career scientists across the world to work on it together. Our conviction is simple. AI advances science, and science advances AI. Neither direction moves as fast as it should, because the open infrastructure both sides depend on — the benchmarks, the verifiers, the record of what is genuinely unsolved — does not yet exist. We are building it in the open, and giving it away. Two projects are launching today. ScienceAccelBench — scientific code, accelerated by AI and verified against the original code's own output. The faster code goes back to the lab that contributed it. https://aitonomy.org/projects/sci-accel ScienceMysteryBench — a finished discovery, rewound to the raw data it came from, and handed to an AI to attempt again. Eight scientific domains, one question each. https://aitonomy.org/projects/science-mystery We are looking for people to build this with. Researchers — join a project. We provide the AI credits and the GPU time; you bring the science. Contributions are credited by name, and lead to a paper on a short timeline. Frontier labs — we would like to work with you. Tell us what evaluations you need, and consider supporting the work with API credits. Sponsors — compute and API credits are what we need most. Everything we build stays open, free, and behind no paywall. Everyone else — a repost genuinely helps us reach the scientists we are looking for. https://aitonomy.org/
media 0 共 1 项
#4 🔗 原推
2026-08-18 08:37:22
Emad
RT @intern: https://x.com/i/article/2089394881353433088
#3 🔗 原推
2026-08-17 19:18:42
Emad
Hit the @grok @bot limit after solid use testing things Using it as a coordinator (gave it my gpt/claude/cursor clis) it feels oddly frustrating versus running out of code model usage I think it’s because everything halts & no warning - should failover to cheap model https://twitter.com/EMostaque/status/2089431551444471855/photo/1
media 0 共 1 项
#2 🔗 原推
2026-08-17 19:17:56
Emad
https://x.com/i/article/2041924211581628416
#1 🔗 原推
2026-04-08 17:03:39