| 来源 | 时间 | 内容(英文原文 / 中文翻译) |
|---|---|---|
| X | 09-09 01:43 | One of the most amazing moments for me in OpenAI history was watching this happen over the past week: 对我而言，OpenAI 历史上最惊人的时刻之一，就是过去一周看着这件事发生。 @sama · 查看原文 |
| X | 09-09 01:23 | We congratulate Levent Alpöge and Tristan Buckmaster on their remarkable mathematical work.
We (the researchers and the agents) did not see any of their work through any means until they released it publicly — in particular, no specific user data was accessed in order to solve 我们祝贺 Levent Alpöge 与 Tristan Buckmaster 的杰出数学工作。在他们公开发布之前，我们（研究人员与智能体）未通过任何途径看到其工作——尤其是，求解该问题时未访问任何特定用户数据。虽不太可能，但我们无法排除源自其产品使用的去标识化数据曾帮助改进模型。不过我们的证明差异显著，在欧拉情形下所证的精确结果也不同（受迫 vs 非受迫）。 @OpenAI · 查看原文 |
| X | 09-09 02:41 | ChatGPT Images 2.5—faster, sharper, smarter, with better tools for creating whatever you can dream of.
- Faster image generation to keep your ideas flowing
- Improved fidelity for more natural, recognizable images
- Consistent details across multiple edits
- Comment-based https://t.co/SjHITKJUSV ChatGPT Images 2.5——更快、更锐利、更聪明，并配备更好的创作工具：图像生成更快、保真度更高、多次编辑细节更一致，以及基于评论的局部修改。 @OpenAI · 查看原文 |
| X | 09-08 22:59 | Subes tu artículo o tesis a ChatGPT para que “corrija” la ortografía.
Tiempo después, un investigador en otro país publica tu trabajo. ChatGPT se lo sugirió cuando utilizó tus datos como entrenamiento. 你把论文或学位论文上传给 ChatGPT“改改错别字”。过一段时间，另一个国家的研究者发表了你的工作。ChatGPT 在用你的数据做训练后，把内容建议给了对方。 @enriquegtejeda · 查看原文 |
| X | 09-09 08:04 | I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below. 我今天从 Anthropic 辞职了。过去三年我在 OpenAI 和 Anthropic 做预训练研究。两家公司都没有负责任地行事。他们正径直冲向可自我改进的超级智能，拿我们的生命做赌注。更多想法见下文。 @hilbertspaess · 查看原文 |
| X | 09-08 21:59 | Bravo to Mistral!
So great to see them killing it, they are an inspiration to everyone as we need more sovereign & open-source AI everywhere around the world.
Excited to keep supporting them on https://t.co/BgguuZx7nO! 祝贺 Mistral！很高兴看到他们大获成功，他们激励所有人——我们需要世界各地有更多主权与开源AI。期待继续在 https://huggingface.co/mistralai 支持他们。 @ClementDelangue · 查看原文 |
| X | 09-09 01:28 | This is a big one.
(And yes, the fights over academic credit and what happened when in the race for the proof need to be resolved, but it is still a big one.) 这是一件大事。（是的，围绕学术署名以及证明竞赛中时间线的争执仍需厘清，但这依然是大事。） @emollick · 查看原文 |
| X | 09-09 05:59 | Terence Tao just posted this really insightful thread which seems to have been written in reaction to the Navier Stokes announcement. Please give it a read. https://t.co/SiuLLyp8jQ 陶哲轩刚发了一串很有洞察力的帖子，看起来是对纳维-斯托克斯宣布的回应。请认真读一下。https://mathstodon.xyz/@tao/117237320796901560 @fchollet · 查看原文 |
| X | 09-09 06:06 | Still more absolute 🔥 from Terence Tao:
“We have now seen that even the rumor of someone working on a problem can trigger a massive amount of AI-powered effort to flatten it before the original research project has time to reach its full potential. The incentives may now be https://t.co/vUrpxRgQA3 陶哲轩又放出绝对高能观点：“我们现在已经看到，哪怕只是有人在做某个问题的传闻，也会触发海量AI算力去抢先把它压平，让原研究项目来不及充分发挥潜力。激励可能正指向：不再向更广泛社群分享任何有前景的研究方向——这将颠倒数百年开放科学传统，并对该领域的长期未来造成严重损害。” @GaryMarcus · 查看原文 |
| X | 09-08 22:03 | We’re launching AlphaGenome Atlas: an AI-powered searchable database mapping the predicted impact of all 9 billion possible single-letter DNA changes.
Here’s how it could help researchers better understand our biology 🧵 https://t.co/SABFZW5SiR 我们推出 AlphaGenome Atlas：一个由AI驱动的可检索数据库，映射全部约90亿种可能的单字母DNA变化的预测影响。以下说明它如何帮助研究者更好理解生物学🧵 @GoogleDeepMind · 查看原文 |
| X | 09-08 23:09 | With AlphaFold we mapped the protein universe - now with AlphaGenome Atlas we’re charting the human genome. It can predict the impact of all 9 billion possible single-letter DNA variants, helping scientists better understand disease. Freely available for academic research: 有了 AlphaFold，我们绘制了蛋白质宇宙；如今借助 AlphaGenome Atlas，我们正在绘制人类基因组。它能预测全部约90亿种单字母DNA变异的影响，帮助科学家更好理解疾病。学术研究可免费使用。 @demishassabis · 查看原文 |
| X | 09-09 01:30 | Announcing a solution to the Navier-Stokes problem, one of the seven Millenium Prize problem.
Represents a major milestone for both AI and mathematics, as well as pointing to an upcoming renaissance in scientific discovery.
Hoping that the knowledge created by this model can 宣布纳维-斯托克斯问题的一个解，这是七个千年奖问题之一。这对人工智能与数学都是重大里程碑，也预示科学发现将迎来复兴。希望该模型产生的知识能帮助学界更好理解非线性偏微分方程——它们被用来建模物理世界的大量现象。 @gdb · 查看原文 |
| X | 09-09 03:45 | Images 2.5 is here.
I don't think it can solve super difficult math problems, but it is really good and we hope you enjoy it.
https://t.co/LdZhGF5AcZ Images 2.5 来了。我不觉得它能解超难的数学题，但确实很好用，希望大家喜欢。https://openai.com/index/introducing-chatgpt-images-2-5/ @sama · 查看原文 |
| X | 09-09 01:20 | We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics.
The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra.
The problem https://t.co/8zol3BPTL4 我们分享纳维-斯托克斯千年奖问题的一个解，这是数学前沿最深刻的问题之一。该证明由一组智能体产出，所用 OpenAI 下一代模型能力显著高于 GPT-6 Astra。该问题关乎纳维-斯托克斯方程所描述的光滑三维流体运动是否会崩溃，约90年来一直未解。 @OpenAI · 查看原文 |
| 来源 | 时间 | 内容(英文原文 / 中文翻译) |
|---|---|---|
| Hacker News 其他 | 09-08 22:05 | LibreOffice breaks download records after declaring it has no AI features Hacker News · 查看原文 |
| 新闻 安全伦理 | 09-08 22:23 | Safety for Whom? Refusing the Right Subset of a Topic, Not the Whole Topic 为谁安全？拒绝某一主题的正确子集，而非整个主题 |
| 新闻 应用落地 | 09-08 22:00 | AlphaGenome Atlas: A predictive map of every possible DNA letter change in the human genome AlphaGenome Atlas：人类基因组中每一个可能DNA字母变化的预测图谱 |
| 新闻 应用落地 | 09-09 01:00 | How GPT-5.6 Sol helps run quantum computing experiments GPT-5.6 Sol 如何助力运行量子计算实验 |
| Hacker News 其他 | 09-08 21:41 | We Must Return to the Office to Use AI in Person Hacker News · 查看原文 |
| Hacker News 应用落地 | 09-08 22:13 | I-have-ADHD: A skill to stop coding agents from burying the answer Hacker News · 查看原文 |
| Hacker News 模型发布 | 09-09 02:37 | ChatGPT Images 2.5 Hacker News · 查看原文 |
| Hacker News 模型发布 | 09-09 03:25 | Muse – Meta’s personal AI agent Hacker News · 查看原文 |
| Hacker News 其他 | 09-08 22:44 | The two Christian saints who are the Buddha Hacker News · 查看原文 |
| Hacker News 安全伦理 | 09-09 03:47 | OpenAI fought dirty on career-making math problem Hacker News · 查看原文 |
| Hacker News 其他 | 09-09 02:26 | Flights cancelled at UK airports due to ATC issue Hacker News · 查看原文 |
| Hacker News 应用落地 | 09-09 00:59 | Show HN: LLM Attention Visualization Hacker News · 查看原文 |
| Hacker News 安全伦理 | 09-09 05:00 | Tao: Open math problems being non-renewably mined by AI Hacker News · 查看原文 |
| Hacker News 模型发布 | 09-09 01:07 | OUI-1: world's first model for Generative UI Hacker News · 查看原文 |
| Hacker News 安全伦理 | 09-09 05:47 | Large language models develop novel social biases through adaptive exploration Hacker News · 查看原文 |
| techcrunch.com 安全伦理 | Tue, 08 Sep 2026 | Chrome has officially switched from a four- to a two-week release schedule, as Google promised earlier this year, with Tuesday’s launch of Chrome 153 on desktop, iOS, and Android. The shift to faster releases is tied to Chrome’s evolving security strategy in the AI era, the company explained. As automated AI tools and community bug reports have pushed up the volume of patches and updates, Google s techcrunch.com · 查看原文 |
| techcrunch.com 模型发布 | Tue, 08 Sep 2026 | On Tuesday, the company introduced Muse, its new personal AI agent that helps consumers with everyday tasks and projects for users in the U.S. To use Muse, consumers will have to trust Meta with more of their personal information than ever before. The AI agent works by connecting to the user’s apps and services that are a part of everyday workflows, like email, calendars, payments, and other thing techcrunch.com · 查看原文 |
| techcrunch.com 资本动向 | Tue, 08 Sep 2026 | Last October, Google said it would bring an Iowa nuclear power plant back from the dead. Now, the facility’s owner, NextEra Energy, has received a $1.9 billion loan from the U.S. Department of Energy to finance the refurbishment. The sizable loan is the second of its kind, suggesting that the Trump administration views revived nuclear power as a key source of electricity for tech companies seeking techcrunch.com · 查看原文 |
| arstechnica.com 算力芯片 | Tue, 08 Sep 2026 | Leading chipmakers Samsung Electronics and Taiwan Semiconductor Manufacturing Co. announced plans to adopt ASML’s latest chipmaking machines in the next several years—and they also joined Intel in agreeing to a crucial technology change that could boost chip production on the new machines by 40 percent. This signifies the semiconductor industry’s broader embrace of high NA EUV photolithography tec arstechnica.com · 查看原文 |
| techcrunch.com 资本动向 | Tue, 08 Sep 2026 | As TechCrunch previously reported, the company, founded just earlier this year by Kofi Asante, builds small nuclear reactors (SMRs) on floating barges, rather than fixed locations, to generate clean energy for ports and nearby infrastructure. This means that energy can be transported to wherever it is needed, an idea Asante hopes can bring clean power to those who need it most. “The Navy is doing techcrunch.com · 查看原文 |
| techcrunch.com 资本动向 | Tue, 08 Sep 2026 | The money could also help clarify its positioning — the company says its goal is not to build a European ChatGPT, and though its models haven’t gone mainstream, it still envisions itself as an AI lab. The funding will help Mistral pursue a subtle shift in strategy that aims to further address concerns in Europe and elsewhere about being too dependent on the United States for tech, especially with techcrunch.com · 查看原文 |
| techcrunch.com 资本动向 | Tue, 08 Sep 2026 | Cognition, the startup developing coding assistant Devin, announced it raised $2 billion at a $48 billion valuation. While the startup didn’t explain how it calculates that run-rate revenue, the metric is usually defined as a month’s top line multiplied by 12. Cursor, another popular coding assistant, was in talks in April to raise capital at a $50 billion valuation before agreeing to sell to Spac techcrunch.com · 查看原文 |
| theverge.com 模型发布 | Tue, 08 Sep 2026 | OpenAI announced ChatGPT Images 2.5 on Tuesday and is adding a new way to tell ChatGPT what you want it to make an image of: by drawing a doodle. With a new feature called Sketch, you can just draw something right inside ChatGPT and then tell ChatGPT how you want it to make an image based on that sketch. You can activate the Sketch feature by typing @Sketch into your chat box, and a window will po theverge.com · 查看原文 |
| theverge.com 其他 | Tue, 08 Sep 2026 | Windows and security engineers have had a busy summer. But this year, the season has been unusually busy for Windows and security engineers, thanks to new AI models that are discovering software vulnerabilities at a rapid pace. Sources tell me that Microsoft will set another patch Tuesday record today, the third in just a few months. Sources familiar with Microsoft’s security work tell me both of theverge.com · 查看原文 |
| techcrunch.com 安全伦理 | Tue, 08 Sep 2026 | He hadn’t been working that day, yet his token usage was climbing. The next day, he disabled everything he had attached to Claude and did not work with it. “In the clearest controlled interval, it increased from 45% to 55% while I performed no work, scheduled Cowork tasks were paused or completed, Dispatch/cloud execution was disabled, and there was no corresponding active local Claude Code task,” techcrunch.com · 查看原文 |
| theverge.com 模型发布 | Tue, 08 Sep 2026 | Meta is making another push to bring artificial intelligence to the masses with Muse, a personal assistant it says can put AI in the hands of virtually anyone. The product is the latest step in a multi-billion-dollar strategy overhaul designed to revitalize the company’s ailing position in the AI race and help it catch up to rivals like OpenAI, Anthropic, and Google. Muse will roll out in the US o theverge.com · 查看原文 |
| arstechnica.com 其他 | Tue, 08 Sep 2026 | # Why this month’s Microsoft patch release is a doozy. Security gnomes are pumping out patches ahead of an expected onslaught of AI-assisted attacks. Microsoft’s patch for September is a doozy, with a record number of roughly 972 vulnerabilities fixed and 112 of them meeting the high critical-severity threshold. Two weeks ago, OpenAI, Anthropic, Amazon Web Services, Google, Microsoft, and 100 comp arstechnica.com · 查看原文 |
| techcrunch.com 安全伦理 | Tue, 08 Sep 2026 | The findings, made in collaboration with Anthropic mathematician Levent Alpöge and using both Codex and Claude AI models, are significant in themselves — but they’re also accompanied by an unusual controversy surrounding OpenAI’s attempts to solve the same problem. “There is another part of this story,” Buckmaster wrote in his statement announcing the proofs, “and one that, honestly, I very much w techcrunch.com · 查看原文 |
| techcrunch.com 资本动向 | Tue, 08 Sep 2026 | Google Cloud and Accenture are working together on a joint unit dedicated to sending engineers into enterprises to help them better adopt Google’s AI tools and services. The new unit, dubbed Accenture Gemini Enterprise Business Group, is Google’s latest foray into the increasingly competitive world of “forward-deployed engineers,” or FDEs. Rivals in the AI race, including OpenAI, Anthropic, Micros techcrunch.com · 查看原文 |
| theverge.com 应用落地 | Tue, 08 Sep 2026 | # Google’s Atlas of the human genome could pave the way for new treatments. DeepMind’s new AI-powered tool can learn patterns between DNA changes and our biology. Google DeepMind has unveiled an AI tool that its scientists claim could help unravel the mysteries of the human genome and transform our understanding of biology, accelerating scientific research and ultimately paving the way for new tre theverge.com · 查看原文 |
| Hacker News 安全伦理 | 09-09 03:43 | Meta Failed to Catch Hundreds of AI Child Abuse Ads Hacker News · 查看原文 |
| Hacker News 安全伦理 | 09-09 08:40 | I Resigned from Anthropic Today Hacker News · 查看原文 |
| Hacker News 其他 | 09-09 00:54 | DOJ Blocked ICE Agent Shooting Charge over Federal Prosecutor's Objections Hacker News · 查看原文 |
| TechEBlog - 应用落地 | 09-09 06:27 | Google DeepMind's AlphaGenome Atlas Places a Score on Every Possible DNA Letter Change - TechEBlog - Google DeepMind的AlphaGenome Atlas为每一个可能的DNA字母变化打分 - TechEBlog - TechEBlog - · 查看原文 |
| IEEE Spectrum 应用落地 | 09-08 22:02 | Google DeepMind Maps 9 Billion Possible DNA Variants - IEEE Spectrum Google DeepMind绘制90亿种可能的DNA变异 - IEEE Spectrum IEEE Spectrum · 查看原文 |
| MarkTechPost 应用落地 | 09-09 03:00 | Google DeepMind Releases AlphaGenome Atlas With Precomputed Molecular Effect Predictions and AVI Scores for 9 Billion Human DNA Variants - MarkTechPost Google DeepMind发布AlphaGenome Atlas，提供90亿个人类DNA变异的预计算分子效应预测与AVI评分 - MarkTechPost MarkTechPost · 查看原文 |
| The Register 应用落地 | 09-09 06:50 | Google DeepMind rises above the AI scrum with genome atlas - The Register Google DeepMind凭借基因组图谱在人工智能混战中脱颖而出 - The Register The Register · 查看原文 |
| SiliconANGLE 应用落地 | 09-09 07:41 | Google DeepMind’s AlphaGenome Atlas maps all 9 billion possible human DNA changes - SiliconANGLE Google DeepMind的AlphaGenome Atlas绘制全部90亿种可能的人类DNA变化 - SiliconANGLE SiliconANGLE · 查看原文 |
| PR Newswire 应用落地 | 09-09 07:27 | Stowers Institute partners with Google DeepMind and leading research institutions to help reveal the regulatory language of the human genome - PR Newswire Stowers研究所与Google DeepMind及领先研究机构合作，助力揭示人类基因组的调控语言 - PR Newswire PR Newswire · 查看原文 |
| Scientific Ameri 应用落地 | 09-08 22:00 | New Google DeepMind atlas could transform our understanding of genetic diseases - Scientific American Google DeepMind新图谱或将改变我们对遗传疾病的理解 - 《科学美国人》 Scientific American · 查看原文 |
| Fortune 应用落地 | 09-08 22:00 | Google DeepMind publishes AI-powered predictions for effect of all 9B mutations to human DNA - Fortune Google DeepMind发布人工智能驱动的预测，涵盖全部90亿个人类DNA突变的影响 - 《财富》 Fortune · 查看原文 |
| blog.google 应用落地 | 09-08 22:08 | AlphaGenome Atlas: a high-resolution map of human DNA - blog.google AlphaGenome Atlas：高分辨率人类DNA图谱 - blog.google blog.google · 查看原文 |
| Google DeepMind 应用落地 | 09-08 22:05 | AlphaGenome Atlas: Molecular predictions for 9 Billion human DNA variants - Google DeepMind AlphaGenome Atlas：90亿个人类DNA变异的分子预测 - Google DeepMind Google DeepMind · 查看原文 |
| aws.amazon.com 算力芯片 | 09-09 00:21 | Benchmarking small LLM inference on SageMaker AI: G7 vs G5 and G6 - aws.amazon.com 在SageMaker AI上对小型大语言模型推理进行基准测试：G7对比G5与G6 - aws.amazon.com aws.amazon.com · 查看原文 |
| shattered.io 模型发布 | 09-09 04:55 | Inception Launches Mercury 2.5: 1,100 Tokens/Sec [2026] - shattered.io Inception发布Mercury 2.5：每秒1100个token [2026] - shattered.io shattered.io · 查看原文 |
| Business Wire 模型发布 | 09-09 00:30 | Inception Launches Mercury 2.5, the Next Tier of Intelligence for Diffusion LLMs - Business Wire Inception发布Mercury 2.5，扩散式大语言模型的下一阶智能 - Business Wire Business Wire · 查看原文 |
| Exclusive.kz 应用落地 | 09-08 21:30 | Large Language Models for Small Language Communities. - Exclusive.kz 面向小语种社群的大语言模型。 - Exclusive.kz Exclusive.kz · 查看原文 |
| aws.amazon.com 模型发布 | 09-09 06:06 | Take on your most ambitious work with GPT-6 Astra on Amazon Bedrock - aws.amazon.com 借助Amazon Bedrock上的GPT-6 Astra应对最具挑战性的工作 - aws.amazon.com aws.amazon.com · 查看原文 |
| About Amazon 应用落地 | 09-09 00:02 | Millions of students can now get access to a free year of Kiro from AWS - About Amazon 数百万学生现可免费使用一年AWS的Kiro - About Amazon About Amazon · 查看原文 |
| BBC 应用落地 | 09-09 04:57 | OpenAI says it cracked 90-year-old maths problem in 88 hours - BBC OpenAI称用88小时破解了存在90年的数学难题 - BBC BBC · 查看原文 |
| The New York Tim 应用落地 | 09-09 05:02 | OpenAI Says It Has Cracked One of Math’s ‘Millennium Problems’ - The New York Times OpenAI称已破解数学“千禧年难题”之一 - 《纽约时报》 The New York Times · 查看原文 |
| WSJ 安全伦理 | 09-09 07:46 | Exclusive | Anthropic Researcher Quits Over ‘Out-of-Control’ AI Fears - WSJ 独家 | Anthropic研究员因“失控”人工智能担忧辞职 - 《华尔街日报》 WSJ · 查看原文 |
| Reuters 监管政策 | 09-09 05:31 | US accuses Chinese AI firms of 'malicious' copying of AI technology - Reuters 美国指责中国人工智能公司“恶意”抄袭人工智能技术 - 路透社 Reuters · 查看原文 |
| 来源 | 市场 | 榜位 | 时间(北京) | 标题(点击查看原文) |
|---|---|---|---|---|
| 知乎 | 综合 | 1 | 09-09 09:17 | 如何看待 Buckmaster 披露 OpenAI 在 NS 方程突破中的学术掠夺与威胁言论？ |
| 知乎 | 综合 | 2 | 09-09 09:17 | 都说 GPT-6-Astra 强，有没有实际点的应用案例？ |
| 贴吧 | 综合 | 2 | 09-09 09:17 | AI冲击短剧,寒冬变坟墓 |
| 知乎 | 综合 | 5 | 09-09 09:17 | 怎么看 DeepSeek Flash 系列 9 月 10 日将再调整定价，除输出外回归 8 月 17 日前价格？ |
| bilibili 热搜 | 综合 | 9 | 09-09 09:17 | 英伟达将推出GB203版本5070 |
| 知乎 | 综合 | 13 | 09-09 09:17 | 怎么看 OpenAI 在 Navier–Stokes 数学难题上取得的进展反而出现争议？ |
| 澎湃新闻 | 综合 | 13 | 09-09 09:17 | 一图世界｜AI抢走白领饭碗，美国打工人面临“学历贬值”？ |
| 百度热搜 | 综合 | 20 | 09-09 09:17 | DeepSeek官宣：调整flash系列定价 |
| bilibili 热搜 | 综合 | 25 | 09-09 09:17 | AI动画 聊斋志异之李伯言 |