Latent SpacePodcast
今日模型发布后市场平静[AINews]今天没发生什么
一周不间断的模型发布后安静的一天a quiet day after a week of nonstop model releases
一周不间断的模型发布后安静的一天a quiet day after a week of nonstop model releases
至少到目前为止，我非常确信人工智能一直在创造净就业机会。这不是我所期望的--尽管我比其他人不那么悲观，但我认为通过这种水平的能力，我们会看到一些impact.it是可能的这个方向继续前进！so far at least, i'm pretty sure AI has been net job-creating.this was not what i expected--although i was much less pessimistic than others, i thought by this level of capability we'd have seen some impact.it is possible this direction keeps going!
whoaTibo：当你1天内的增长量超过过去两周的总和时，你就知道这是一个很好的模型。从https://chatgpt.com/download/安装桌面应用程序并尝试一下！whoaTibo: You know it's a good model when you grow more in 1 day than last two weeks combined.Install the desktop app from https://chatgpt.com/download/ and give it a go!
“医生发现GPT-5.6回复中的缺陷比医生书面回复少。“Karan Singhal：GPT-5.6是健康事业向前迈出的一大步，无论是在前沿还是在成本方面。这些模型推动了每美元性能的前沿，为所有人带来最好的健康情报。️最小的var"physicians found fewer flaws in GPT-5.6 responses than physician-written responses."Karan Singhal: ♥️ GPT-5.6 is a major step forward for health, both at the frontier and at cost.These models push the frontier of performance per dollar, bringing the best health intelligence to all. The smallest var
我们办公室的新海报显示了Codex使用限制的重置如何影响我们的部分系统。数据是美丽的。New poster for our office that shows how resets of the Codex usage limits affects part of our systems. Data is beautiful.
当您1天内的增长量比过去两周的总和还要多时，您就知道这是一个很好的模型。从https://chatgpt.com/download/安装桌面应用程序并尝试一下！You know it's a good model when you grow more in 1 day than last two weeks combined.Install the desktop app from https://chatgpt.com/download/ and give it a go!
如果您以前使用过Codex，请做两件事1。继续使用Codex。这是给你的2。告诉您的朋友和家人有关ChatGPT Work的信息您一直想向Codex展示，但他们技术含量不高，也不是每天都使用代码，并告诉我进展如何If you were using Codex before, two things to do1. Keep using Codex. It is for you2. Tell your friends and family about ChatGPT Work who you always wanted to show Codex but they’re not as technical and don’t work with code every dayand tell me how it goes
这超级酷，但我想知道所有这些基准是如何工作的？如何衡量设计品味？Arena.ai：令人兴奋的消息：@OpenAI的GPT-5.6-sol现已在Code Arena：Frontend中并列第一，与Claude Fable 5相匹配！这标志着OpenAI模型首次登上Code Arena的榜首，恶魔This is super cool but I wonder how all these benchmarks work? How do you benchmark design taste?Arena.ai: Exciting news: @OpenAI’s GPT-5.6-sol is now joint #1 in the Code Arena: Frontend, matching Claude Fable 5!This marks the first time an OpenAI model has reached the top spot in Code Arena, demon
本周末我的新项目流程：->使用Fable构建带有设计指南的plan.html->使用Claude Design制作组件+屏幕->获取GPT 5.6来构建My process for a new project this weekend:-> Use Fable to build plan.html with design guidelines-> Use Claude Design to make components + screens-> Get gpt 5.6 to build
RT Matt Van Horseman Printing Press现在可以打印自学习CLIs从现在开始，它打印的每个CLI都会留下自己的笔记，因此您的代理每次运行都会变得更聪明。视频中的真实基准：同样的问题从10个电话变成了2个电话。在世界杯之前打印了ESPN CLI...它学会了自己RT Matt Van HornThe Printing Press now prints SELF-LEARNING CLIs Every CLI it prints from now on leaves itself notes, so your agent gets smarter every run. Real benchmark in the video: same question went from 10 calls to 2 calls. Printed an ESPN CLI before the World Cup existed... it taught itself h
汽车公司善于命名事物。每个人都知道我说“同意”是什么意思。他们知道这是一辆本田。他们可能会问“哪一年？”或“GX？“这是4个维度（品牌、型号、年份、修剪），每个维度对功能和质量有不同的影响。南宇：简化事物仍然是一个困难 Car companies are good at naming things. Everyone knows what I mean when I say "Accord". They know it's a Honda. They might follow up with "what year?" or "LX?"That's 4 dimensions (make, model, year, trim) each with different implications on features and quality.Nan Yu: Naming things remains a hard
RT Jason SaaStr.Ai Lemkin我不知道你可以在Claude内部使用Replit LCP。我今天打开/连接了它们，Claude和我们的Replit代理立即开始谈论、辩论，并喋喋不休地谈论我们正在发货的新产品中的错误和问题。我实际上跟不上👾RT Jason ✨👾SaaStr.Ai✨ LemkinI had no idea you could use Replit MCP inside of Claude.I turned it on / connected them today, and immediately, Claude and our Replit agent started talking, debating, and going on and on about bugs and issues in a new product we’re shipping.I actually couldn’t keep up wi
刚刚为我们的销售团队审查了第二季度业绩，我们实现了目标的178%！这个目标已经雄心勃勃，特别是因为去年我们只有几名销售人员。如今，收入团队已接近100名，我们将政府、银行和增长最快的科技公司视为客户。Just reviewed Q2 results for our sales team and we hit 178% of our goal!That goal was already absurdly ambitious, especially since we had only a couple of sales people last year.Today revenue team is closing on 100 and we count governments, banks, and fastest growing tech companies as customers.
RT Jahid我刚刚取消了我的@NotionHQ每月20美元的商业帐户，因为我使用@Replit构建了整个类似于概念的应用程序供我们的团队使用。它工作得很好。它只花了我不到100美元。拥有基础设施。RT JahidI just canceled my @NotionHQ $20/month business account as I built entire notion-like app using @Replit for our team to use. It works just fine. It just cost me under $100. Own the infrastructure.
如果您正在考虑企业中的应用人工智能，这是一篇很棒的文章。这篇文章的标题是关于公司从人工智能中获得的巨大好处，但最深刻的收获是关于人工智能转型在组织中的样子。它从根本上讲是关于改变底层工作流程或bThis is a great post if youre thinking about applied AI in the enterprise. The headline of this post is about what companies have huge upside from AI, but the deepest nuggets are about what AI transformation looks like in an organization.It’s fundamentally about changing the underlying workflow or b
Defund the grifters who created the SF Doom LoopKane 謝凱堯: In 2023 @TheCoalitionSF, an unaudited NGO funded by @sfgov that was using taxes to give free tents to drug traffickers, asked the courts to block towing vehicles with unpaid tickets.Mayor @DanielLurie should defund them and require audits for
RT Varunram Ganesh在这个平台上，有一种极其令人厌恶（不幸的是非常受欢迎）的参与诱饵方式，其中涉及WC，其中包括评论WC、关于WC的随机虚假阴谋论，以及在这种情况下直接对YCI撒谎，如果你有RT Varunram GaneshThere's an extremely distasteful (and unfortunately very popular) way to engagement bait on this platform that involves YC that includes commenting on YC, random false conspiracy theories about YC, and in this case straight up lying about YCI don't get why you'd do that, if you hav
让经纪人想要的东西迈克尔：thinking machines实际上是一个非常有趣的公司仪表板慢到无法使用，但你现在可以通过删除api键并告诉fable训练一个模型来训练模型，所以这并不重要它实际上是通过api训练模型，但比oa更灵活。Make something agents wantMichael: thinking machines is actually such a funny companythe dashboard is slow enough to be unusable but you can now train a model by dropping the api key and telling fable to train one so it doesn’t matterit’s literally train models by api, but more flexible than what oa
Ok Fable和Sol在几个小时内就为我提供了一个基于后台和网络应用程序的完全可用的iOS应用程序。我们生活在一个多么疯狂的世界啊Nikunj Kothari：我认为人们仍然睡在实时语音驱动的消费者体验上。发布我最新的针对孩子们的副业项目：Upon A Voice Ok Fable and Sol got me a fully working iOS app in a few hours based on the backend and the web app.. what an insane world we live in 🤯Nikunj Kothari: I think people are still sleeping on realtime voice driven consumer experiences..Releasing my latest side project geared towards kids: Upon A Voice
在向您提供架构评估之前，让我验证几个承载事实。Let me verify a couple of load-bearing facts before I give you the architecture assessment.
RT Sam Altman有很多基准表明5.6 sol是目前世界上最好的模型，但最可靠的判断方式是埃隆再次痴迷于我RT Sam Altmanthere are a lot of benchmarks that suggest 5.6 sol is the best model in the world right now, but the most reliable way to tell is that elon is obsessed with me again
似乎很糟糕Gerrit De Vynck：有关一名苹果员工的重大指控，该员工于2026年1月加入OpenAI，持有笔记本电脑，在苹果系统中发现漏洞并在几周内窃取文件。🦭这与安东尼·莱万多夫斯基（Anthony Levandowski）在2018年被谷歌指控的行为类似seems badGerrit De Vynck 🦭: Major allegations about an Apple employee who joined OpenAI in Jan 2026, kept a laptop, found an exploit in Apple's systems and stole files over a multi-week time. This is similar to what Anthony Levandowski was accused of doing by Google back in 2018
人工智能中的一切都发展得如此之快。我们将到达这里：-运行云中的所有代理-选择任何模型（前沿，OSS，中国，美国）-选择任何工具-拥有完整的跟踪-拥有递归改进循环我知道它会在那里，但它会已经发生吗？Everything in AI moves so fast.We are going to get here:- Run all the agents in the cloud- Choose any model (frontier, OSS, Chinese, American)- Choose any harness- Have full tracing- Have recursive improvement loopsI know it will be there but can it happen already?
RT Amjad Masad刚刚回顾了我们销售团队的第二季度业绩，我们达到了目标的178%！这个目标已经是非常雄心勃勃的了，尤其是去年我们只有几个销售人员。今天，我们的收入团队接近100人，我们把政府、银行和增长最快的科技公司都算在内， RT Amjad MasadJust reviewed Q2 results for our sales team and we hit 178% of our goal!That goal was already absurdly ambitious, especially since we had only a couple of sales people last year.Today revenue team is closing on 100 and we count governments, banks, and fastest growing tech companies as
对于OpenAI来说是重要的一天。A big day for OpenAI.
RT TiboHello美丽的人！我们已经重置了Codex和ChatGPT Work的使用限制。另一个将在当天晚些时候到来。高兴吧。现在我引起了你的注意，一个关于ChatGPT Work的快速更新，Codex和我们昨天分享的所有更新。我们花了过去24小时阅读反馈，lRT TiboHello beautiful people! We have reset usage limits across Codex and ChatGPT Work. And another one will come later in the day. Rejoice.Now that I have your attention, a quick update on ChatGPT Work, Codex and all the updates we shared yesterday.We’ve spent the last 24 hours reading feedback, l
GPT-5.6现在是Microsoft 365中的首选型号Copilothttps：openai.com/index/gpt-5-6-preferred-model-microsoft-365-copilot/GPT-5.6 is now the preferred model in Microsoft 365 Copilothttps://openai.com/index/gpt-5-6-preferred-model-microsoft-365-copilot/
我们很高兴看到人们如此喜欢5.6索尔！makes us happy to see people love 5.6 sol so much!
RT AI工程师“我认为让人们理解代码是如何工作的很重要。“Geoffrey Litt的Design Eng主题演讲现已上线：https://www.youtube.com/watch? v= WkBPX-oDMnAT感谢您@NotionHQ对他令人难以置信的工作的支持，并感谢Geoff为我们的设计工程师启动WF 26！杰弗RT AI Engineer"I think it's important for people to understand how code works."Geoffrey Litt's Design Eng track keynote is live now:https://www.youtube.com/watch?v=WkBPX-oDMnAThank you @NotionHQ for supporting his incredible work and thank you Geoff for kicking off WF26 for our Design Engineers!Geof
RT Alex Volkov @ AI Engineerhttps：www.youtube.com/watch? v= ZpK 5 PWX 2 YRM AI工程师在2026年还应该读代码吗？这显然是一个分裂的采取，有像西奥，桥本，普里米根，甚至叔叔鲍勃插话。我预测这两个月前，并命名为ZL连续体。我的tRT Alex Volkov @ AI Engineerhttps://www.youtube.com/watch?v=ZpK5PWX2YRMShould AI Engineers read code anymore in 2026? This, apparently is a divisive take, that has folks like Theo, Hashimoto, Primeagen and even uncle bob chiming in.I predicted this two months ago, and named it the ZL Continuum. My t
RT马修·伯曼。@ swyx和我在AIE坐下来谈论AGI、前沿实验室策略、芯片、寓言、政府和人工智能存在主义RT Matthew Berman.@swyx and I sat down at AIE to talk about AGI, frontier lab strategy, chips, Fable, government, and AI existentialism
@GeminiApp可信测试者计划的更新：我们已经加入了第一批队列（如下），并发出了一些挑战#1的邀请：测试我们的macOS应用程序中即将推出的未发布功能。更多邀请和挑战即将推出，如果您匹配，请密切关注您的收件箱！乔什·伍德沃德：想要不受欢迎的，Update on @GeminiApp Trusted Testers Program:We've onboarded the 1st cohort (below) and sent out some invites for Challenge #1: testing unreleased features coming soon in our macOS app.More invites and challenges dropping soon, keep an eye on your inbox if you're matched!Josh Woodward: Want unreleas
印度KBC x双子座！谷歌印度：梦想成为KBC的热门席位吗？让我们带您去那里Google Gemini正在与KBC和@SonyTV合作，作为您的个人知识伙伴从注册到最终的padhaav，只需一个提示即可获得答案。Toh chaliye shuru krte hai http：//goo。KBC x Gemini in India!Google India: Dream of making it to the KBC hot seat? Let's get you there 🤝Google Gemini is teaming up with KBC and @SonyTV as your personal knowledge companion ✨From registrations to the final padhaav, your answers are just a prompt away.Toh chaliye shuru krte hai http://goo.
美丽的人们好！我们已经重置了Codex和ChatGPT Work的使用限制。另一个将在当天晚些时候到来。高兴。现在我引起了您的注意，请快速更新ChatGPT Work、Codex以及我们昨天分享的所有更新。我们花了24小时阅读反馈，寻找 Hello beautiful people! We have reset usage limits across Codex and ChatGPT Work. And another one will come later in the day. Rejoice.Now that I have your attention, a quick update on ChatGPT Work, Codex and all the updates we shared yesterday.We’ve spent the last 24 hours reading feedback, looking
这实际上不是Cerebras上GPT-5.6 Sol的版本。只是正常/快速模式。在750+ TPS下运行则完全是另一回事。Chamar：在Blender中750 Token/s上GPT-5.6 Sol。没有加速。️天哪。这比查看基准更令人印象深刻This actually not the version of GPT-5.6 Sol on Cerebras. Just normal /fast mode. Running it at 750+ TPS is a whole other thing.Chubby♨️: GPT-5.6 Sol on 750 Token/s in Blender.Not sped up. Holy moly. That’s so much more impressive than looking at benchmarks
为了庆祝GPT-5.6 Sol的推出，我们将在接下来的24小时内再次重置ChatGPT Work和Codex的速率限制（两次）。我们希望您有时间真正尝试雄心勃勃的任务并掌握窍门。探索快乐！To celebrate the launch of GPT-5.6 Sol, we will reset the rate limits again (twice) across ChatGPT Work and Codex over the next 24 hours.We want you to have the time to truly try ambitious tasks and get the hang of it. Happy exploring!
比利时踢得很好，很遗憾看到失利。Belgium played well, sorry to see the loss.
为什么英国播音员不做“GOOOOOOAAL”的事情。这应该是工作的必要部分哈哈Why doesn't the English announcers do the "GOOOOOOAAL" thing. This should be a required part of the job lol
需要指出的一件事--我根本没有考虑过在5.5高速模式下的使用。感觉基本上是无限的。5.6 Sol高速模式使用起来很快。要切换5.6 Sol medium no fast模式，看看它在浏览器使用和其他任务上的表现如何。Peter Yang：OpenAI刚刚制作了GPT-5.6 avOne thing to point out - I never thought about usage at all on 5.5 high fast mode. It felt basically unlimited.5.6 Sol high fast mode burns through usage pretty fast. Going to switch 5.6 Sol medium no fast mode to see how it does on browser use and other tasks.Peter Yang: OpenAI just made GPT-5.6 av
狗吐-当你狗食你的产品太他妈的硬，并使用它的方式是完全脱离你的实际客户。Dog-barfing — when you dogfood your product too damn hard and use it in ways that are completely divorced from your actual customers.
ChatGPT Codex Adeptus Astartes
Claude Code现在可以打开桌面应用程序内的任何网站。Claude可以使用您的制作应用程序，打开它向您发送的链接，查看Twitter，甚至观看FIFA世界杯！ClaudeDevs：桌面上的Claude Code现在有一个应用内浏览器。Claude可以调出文档、设计或任何其他网站。它可以阅读，Claude Code can now open any website inside the desktop app.Claude can use your production app, open the links it sends you, check Twitter, and even watch the FIFA World Cup!ClaudeDevs: Claude Code on desktop now has an in-app browser.Claude can pull up docs, designs, or any other site. It can read,
RT兰斯·马丁http：x.com/i/article/2075610072118702081RT Lance Martinhttp://x.com/i/article/2075610072118702081
桌面上的RT ClaudeDevsClaude Code现在有一个应用内浏览器。Claude可以调出文档、设计或任何其他网站。它可以像与本地开发服务器一样读取、点击和交互。它是沙箱式的且可配置的：您选择会话是否持续存在。RT ClaudeDevsClaude Code on desktop now has an in-app browser.Claude can pull up docs, designs, or any other site. It can read, click through, and interact the same way it does with your local dev servers.It's sandboxed and configurable: you choose whether sessions persist.
尝试过Claude Design的RT nate parrott人们-体验如何？什么效果好，什么可以更好？RT nate parrottfolks who’ve tried Claude Design — how was the experience? what works well and what could be better?
RT TIME人工智能行业一直痴迷于“代币最大化”。“在@RaiseSummit上，@Replit认为是时候掉头了，而是提出了“结果最大化”的理由。RT TIMEThe AI industry has been obsessed with "token-maxxing." At @RaiseSummit, @Replit argues it's time for a U-turn, making the case for "outcome-maxxing" instead.
RT Nadia I看看你可以在@Replit画布中做什么！向@Noni_Shehnoor大喊，感谢他们教我如何创建区域编辑器并利用画布来实现精确性。🤯🔬We have an episode on this at http://Classroom2Code.com if youre curious :)RT Nadia ICheck out what you can do in @Replit canvas! Shout out to @Noni_Shehnoor for teaching me how to create a region editor and utilize the canvas for precision. 🤯🔬We have an episode on this at http://Classroom2Code.com if youre curious :)
RT Manny Bernabe今天的有趣节目：社区简介展示！由一些令人惊叹的建筑商回顾他们的个人资料和他们正在建造的东西：Nadia Iftekhar（@Niftek）和Noni Ansari（@Noni_Shehnoor）、Victoria Cavazos和Marcel Haas（@marcelhaasIO）。另外：·我们正在宣布所有25名获奖者RT Manny BernabeFun show today: Community Profile Showcase! 🤩Featuring somme amazing builders walking through their profiles and what they're building: Nadia Iftekhar (@NIftek ) & Noni Ansari (@Noni_Shehnoor ), Victoria Cavazos, and Marcel Haas (@marcelhaasIO ).Plus:• We're announcing all 25 winner
我喜欢@shadCN成功地为代理和人类构建了最好的UI基础设施。使用新的typeup配置器真是太令人满意了。shadcn：介绍shadcn/typeup。您知道如何渲染markdown并恢复纯、无样式的HTML吗？标题、段落、列表、表格。 I love that @shadcn has managed to build the best UI infrastructure for agents and for humans. It's so so satisfying to play with the new typeset configurator.shadcn: Introducing shadcn/typeset. You know how you render markdown and get back plain, unstyled HTML? Headings, paragraphs, lists, tables.
深入了解如何在@v0中使用自己的设计系统生成设计、原型、前端和全栈应用。随着模型不断变得更快，v0变得更加神奇。v0：http://x.com/i/article/2075567583298977792Deep dive into how you can generate designs, prototypes, frontends, and full stack apps with your own design system in @v0. As models continue to get faster, v0 gets more magical.v0: http://x.com/i/article/2075567583298977792
我现在对人工智能有着不健康的痴迷的原因是，我的整个职业生涯基本上都在研究一个问题：如何提高企业中内容的价值。如何保护它，如何合作，如何治理它，以及如何将它集成到各个领域The reason I have an unhealthy obsession with AI right now is because I've spent my entire professional life on essentially one problem: how do you increase the value of content in the enterprise. How do you secure it, how do you collaborate on it, how do you govern it, and how to integrate it acros
当人工智能模型在企业之间共享并为该行业的智能带来如此多的信息时，需要思考未来世界竞争优势的一些动力。这将成为未来十年公司和整个经济的核心问题Great post on some of the dynamics to think through for the future competitive advantage in world when AI models are shared amongst firms and packing so much for the intelligence of that industry.This is going to become a core question for companies and the economy broadly over the next decade and b
开重LLM会成为下一个晶体管吗？Will open weight LLMs be the next transistor?
Insforge就是当自治代理只想编码而不是处理奇怪的API和云服务器和服务的入职时所发生的事情，这些服务是为缓慢而缓慢的人类提供的Hang Huang：我们现在已经启动并服务了超过40，000个项目@insforge！几个月前，我们还在努力Insforge is what happens when autonomous agents just want to code and not deal with weird APIs and onboarding for cloud servers and services that are made for squishy slow humansHang Huang: We've now launched and served 40,000+ total projects @insforge!A few months ago, we were still struggling to e
萎缩既愚蠢又无意中有趣Armand Domalewski：我参加了一个萎缩活动，他们要求每个人带一份礼物并接受一份礼物，这不可避免地导致每个人都带了他们最不想保留的东西，而几乎没有人拿走其他人不想要的东西Degrowth is both stupid and unintentionally funnyArmand Domalewski: i attended a degrowth event where they asked everyone to bring a gift and take a gift, which inevitably lead to everyone bringing the thing they least wanted to keep and almost no one taking the things that other people did not want
到目前为止，法国队最令人惊讶的是，尽管奥利塞有助攻天才，但他还没有进球。The most surprising thing about the France team so far is that Olise, for all his assist genius, hasn’t scored yet.
法国队和挪威队有很好的机会在决赛中再次出场，这将是一个非常罕见的事件（自从两支球队在同一届男子世界杯的小组赛和决赛中相遇以来，已经过去了60多年），也很吸引人（哈兰的改变有多真实？）Decent chance France and Norway could play again in the final, which would be an extremely rare occurrence (been over 60 years since two teams met in both the group stage and the final of the same men's World Cup) and fascinating (how truly game-changing is Haaland?)
我认为人们仍然睡在实时语音驱动的消费者体验上。发布我最新的面向孩子的边项目：在一个声音的基础上你可以看到我的女儿在下面的视频与它互动，但亮点如下：-生成一个“活”的故事书基于谁你的孩子 I think people are still sleeping on realtime voice driven consumer experiences..Releasing my latest side project geared towards kids: Upon A Voice ✨You can watch my daughter interact with it in the video below but the highlights are as follows:- generates a "live" storybook based on who your child
对于创始人和风险投资者来说，撒了酱，但有两个普遍的真理。1)如果你看到他们在X上发出很大的噪音，他们肯定是在筹款。尽管他们不是。2)每一个抛出的事件都是一个来源事件（风投的创始人，创始人的人才）。那个友好的扑克之夜Spilling the sauce but two universal truths when it comes to founders & VCs..1) if you see them make a lot of noise on X, they ARE definitely fundraising.. even though wink wink they are not. 2) every event thrown is a sourcing event (founders for VCs, talent for founders). that friendly poker night
我向妻子解释本周模特发布情况..“宝贝听着。这很重要。GPT-5.6分为三种口味。索尔、泰拉和卢娜。索尔是聪明的。它必须在释放之前清除政府的安全隐患，例如获得假释。Luna是最便宜的。Terra存在。Grok 4.5发布 Me explaining this week’s model release situation to my wife.."babe listen. this matters. GPT-5.6 dropped in three flavors. Sol, Terra, and Luna. Sol is the smart one. It had to clear government safety evals before release, like getting paroled. Luna is the cheap one. Terra exists.Grok 4.5 launched
RT Gerspel Orosz当您意识到X >> Threads适用于任何与技术相关的内容时，马克·扎克伯格（Mark Zuckerberg）在这里和Threads上发布了完全相同的发布帖子（我从来不明白Threads如何声称比X更多的用户。感觉就像一个科技鬼城）RT Gergely OroszThe moment you realize that X >> Threads for anything tech-related in reach / engagementMark Zuckerberg posted the exact same launch post here and on Threads(I never understood how Threads claims more users than X. Feels like a tech ghost town there)
RT Aaron Stannard违背我的意愿，我被迫了解Bun / Zig戏剧。我的理解是这样的：- Bun == OpenCode和Claude Code崩溃的东西- Rust ==毛茸茸的手工编程语言- Zig ==对于认为Rust太主要的开发人员来说，甚至更多手工编程语言RT Aaron StannardAgainst my will, I have been forced to learn about the Bun / Zig drama. Here's my understanding:- Bun == thing that crashes OpenCode & Claude Code- Rust == artisanal programming language for furries- Zig == even more artisanal programming language for devs who think Rust is too main
他们这么做了SClaudeDevs：桌面上的Claude Code现在有一个应用内浏览器。Claude可以调出文档、设计或任何其他网站。它可以像与本地开发服务器一样读取、点击和交互。它是沙箱式的且可配置的：您选择会话是否持续存在。THEY DID IT FOLKSClaudeDevs: Claude Code on desktop now has an in-app browser.Claude can pull up docs, designs, or any other site. It can read, click through, and interact the same way it does with your local dev servers.It's sandboxed and configurable: you choose whether sessions persist.
RT chekosic，如果这是真的，但对我来说@Danshipper促成了这一切lolClaudeDevs：桌面上的Claude Code现在有一个应用内浏览器。Claude可以调出文档、设计或任何其他网站。它可以像与本地开发服务器一样读取、点击和交互。它是沙箱和配置的RT chekosidc if it's true but to me @danshipper made this happen lolClaudeDevs: Claude Code on desktop now has an in-app browser.Claude can pull up docs, designs, or any other site. It can read, click through, and interact the same way it does with your local dev servers.It's sandboxed and configura
如果你关心人工智能的潜在危害，这无疑是最大的风险领域，而不是科幻小说的失控安东尼奥·朱利奇：在尼日利亚东北部的一家酒店房间里，我打开了一个领先的人工智能聊天机器人，将我的笔记本电脑转向一名前博科圣地指挥官，并询问他是否使用过它。他点点头。“你输入if you care about potential harm from AI this undoubtedly the biggest risk area, not a sci fi runaway scenarioAntonia Juelich: In a hotel room in northeast Nigeria, I opened a leading AI chatbot, turned my laptop toward a former Boko Haram commander, and asked if he'd used it. He nodded."You type in
Cognition和Cursor都发布了基于OSS模型的近前沿编码模型。这很有趣.但更重要的是，我们将在未来6个月内看到所有高价值领域的特定领域模型。长期开放+后训练。Cognition and Cursor have both released near-frontier coding models based on OSS models.That is interesting...but more so as an indication of the fact that we will see domain-specific models within the next 6 months for all the high-value domains.Long Open + post-trained.
OpenAI的最新旗舰型号今天上午正式上市，有三种尺寸：Luna、Terra和Sol（从最小到最大）。新型号每100万个输入/输出代币的定价为Luna 1美元/6美元、Terra 2.50美元/15美元、Sol 5美元/30美元。相比之下，克劳德作品系列是5美元/25美元， OpenAI's latest flagship model hit general availability this morning, and comes in three sizes: Luna, Terra, and Sol (from smallest to largest). The new models are priced per 1M input/output tokens as Luna $1/$6, Terra $2.50/$15, Sol $5/$30. For comparison, the Claude Opus series are $5/$25 and the
重写只需11天即可完成，一个小团队需要一年的时间才能完成，代币价值16.5万美元。另外：LLM编码战争升温、来自朝鲜的人工智能造假者在招聘时仍然是一个问题等等A rewrite done in 11 days that would have taken a small team a year to complete, for $165K in tokens. Also: coding LLM “wars” heat up, AI fakers from North Korea still a problem when hiring, and more
SpaceXAI的发展速度仍然比地球上任何其他前沿实验室都快。SpaceXAI continues to move faster than any other frontier lab on earth.
当格伦·福格尔（Glenn Fogel）于2000年加入Priceline时，该业务价值数亿美元。一周后，纳斯达克指数见顶，最终导致其股价跌至每股1美元。但25年过去了，Booking Holdings的规模已超过1000倍，成为价值超过1000亿美元的全球旅行公司When Glenn Fogel joined Priceline in 2000, the business was worth a few hundred million dollars. One week later, the Nasdaq peaked, eventually sending its stock down to a dollar a share. But over 25 years later, Booking Holdings has scaled over 1000x into an over $100 billion dollar global travel be
Jürgen Schmidhuber博士是一位著名科学家和人工智能研究员，被广泛认为是该领域的先驱之一，他通过实验室的工作提出了当今变压器、LSTM和回归自我改进背后的关键想法。他认为真正的AGI仍然受到物理硬件的限制，即 Dr. Jürgen Schmidhuber, a renowned scientist and AI researcher widely regarded as one of the pioneers in the field, originated key ideas behind today's transformers, LSTMs, and recursive self-improvement through his lab's work. He argues that true AGI remains bottlenecked by physical hardware, that
互联网准备好让人工智能代理接管我们的钱包并经营自己的业务了吗？在The MAD播客的本期节目中，Stripe的Emily Sands揭示了代理商业如何从假设的概念迅速转变为部署的金融基础设施。从对抗崛起的存在主义Is the internet ready for AI agents to take over our wallets and run their own businesses? In this episode of The MAD Podcast, Stripe's Emily Sands reveals how agentic commerce is rapidly shifting from a hypothetical concept to deployed financial infrastructure. From combating the rising existential
我对此真的很难过，非常感谢菲吉为Openai所做的一切，甚至感谢她的友谊和她作为一个人的身份。我们都祝愿她早日康复。这太糟糕了。Fidji Simo：今天，我与OpenAI团队分享了我决定辞去我的全职职位，i am really sad about this and very grateful for all fidji has done for openai, and even grateful for her friendship and who she is as a person.we all wish her the best for a speedy recovery. this sucks.Fidji Simo: Today, I shared with the OpenAI team that I have decided to leave my full-time role a
看看这个！您可以完成一些令人惊叹的事情。Codex是我们新工作产品的核心，也是它如此出色的原因。Codex不会去任何地方。OpenAI：引入ChatGPT Work，这是ChatGPT中的一种新代理，由Codex和GPT-5.6提供支持。它可以在您的应用程序和文件中采取行动，并与项目保持一致check this out! you can get some amazing things done.codex is the core of our new work product and what makes it so good. codex is not going anywhere.OpenAI: Introducing ChatGPT Work, a new agent in ChatGPT powered by Codex and GPT-5.6.It can take action across your apps and files, stay with a proje
我们已经听到了企业对人工智能成本的担忧，5.6 sol对于每项任务的美元来说是一个巨大的进步，terra和luna也是如此we have heard enterprises on their concerns about AI costs, and 5.6 sol is a huge step forward for dollars-per-task, as are terra and luna
RT Tejal PatwardhanGPT-5.6 sol训练后的luna！RT Tejal PatwardhanGPT-5.6 sol post-trained luna!
你们这些家伙在@greptile上小丑，但事实证明他们只是@openai的小丑，所以难度要大得多https://x.com/JangLawrenceK/status/2075204015890325703Lawrence Jang：伙计们，我刚刚取消了我的Claude计划，我不知道发生了什么Re u guys were clowning on @greptile but turns out they were just the inspo for @openai to go so, so much harder https://x.com/JangLawrenceK/status/2075204015890325703Lawrence Jang: guys I just cancelled my Claude plan I don’t know what happened
RT AI工程师向@OpenAI祝贺备受期待的GPT 5.6 Sol、Terra和Luna发布！🆕https：//youtu.be/pMGgiOb18tc以下是@romainhuet、@embirico和@steipete，讲述了为什么这是人工智能工程黄金时代的曙光：RT AI Engineer🆕 Congrats to @OpenAI on the highly anticipated launch of GPT 5.6 Sol, Terra, and Luna!https://youtu.be/pMggiOb18tcHere's @romainhuet, @embirico, and @steipete on why this is the dawn of the Golden Age of AI Engineering:
RT Akshat Bubna与@swyx的有趣对话，讲述我们为真正的弹性推理、沙箱等构建云的旅程。当然，还有我们如何改进Modal的开发体验，使其对代理更好。潜在。空间：Modal的代理原生云：DX-AX、沙箱、弹性推理和100，000 roRT Akshat BubnaFun conversation with @swyx on our journey building the cloud for true elastic inference, sandboxes, and more. And of course, how we're evolving Modal's dev experience to be better for agents.Latent.Space: Modal's Agent-Native Cloud: DX→AX, sandboxes, elastic inference, and 100,000 ro
感谢前12小时内收到了1，400多条回复！我昨晚读完了所有这些，一夜之间又收到了300多篇，我今天会补上。以下是迄今为止您反馈的前10名堆栈排名，以及我们对每个堆栈的看法：1）使Google Workshop集成更可靠地工作：这 Thanks to the 1,400+ replies in the first 12 hours! I read all of them last night, and another 300+ came in overnight that I’ll catch up on today. Here’s the Top 10 stack rank from your feedback so far, along with where we stand on each:1) Make Google Workspace integrations work more reliably: This
享受ChatGPT Work和Codex的使用限制的完全重置。在接下来的一个小时里，rajanagarwal刚刚加入，致力于模型研究和推动编码能力。你应该感谢他今天按下了按钮。Enjoy a full reset of your usage limits for ChatGPT Work and Codex. Propagating in the next hour.@_rajanagarwal just joined to work on model research and push on coding capabilities. You can thank him for pressing the button today.
世界的建设者。这是你们的时刻。OpenAI开发人员：OpenAI构建周即将到来。挑战将于7月13日开始。本周，与世界各地的建筑商一起参加现场会议和社区活动。注册是开放的。将想法放在您的积压中。使用Codex构建它。Builders of the world. This is your time.OpenAI Developers: OpenAI Build Week is here.The challenge starts July 13. Across the week, join live sessions and community events with builders around the world.Registration is open.Bring the idea sitting in your backlog.Build it with Codex.
如今，OpenAI产品和工程团队正在蓬勃发展，在GPT 5.6 Sol/Terra/Luna的发布、ChatGPT Work、新的ChatGPT桌面应用程序、托管网站和一百万件其他事情中，有很多事情需要协调和保持运行.以及昨天刚刚推出的新ChatGPT Voice。WToday, the OpenAI product and engineering team is humming, there is so much to coordinate and keep running across the release of GPT 5.6 Sol/Terra/Luna, ChatGPT Work, the new ChatGPT desktop app, hosted sites and a million other things... together with the new ChatGPT Voice just launched yesterday.W
对OpenAI发布的一些赞扬和反馈：1.与任何其他实验室相比，OpenAI更有机会使与代理的合作成为主流。ChatGPT具有图像、实时语音、浏览器/计算机使用以及所有您最喜欢的应用程序的插件，这是最接近与超级有能力的同事合作的事情 Some praise and feedback about OpenAI's launches:1. More than any other lab, OpenAI has the opportunity to make working with agents mainstream. ChatGPT with images, live voice, browser/computer use, and plugins for all your favorite apps is the closest thing to working with a super-capable coworker
呃法国太拥挤了Ugh France is way too stacked
顺便说一句，上面的视频仅使用GPT 5.6，使用ffmpeg和浏览器+电脑制作。GPT玩了游戏并录制了它。Re Btw the video above was made only using GPT 5.6 using ffmpeg and browser + computer use.GPT played the game and recorded it.
RT Karri Saarinen当一家公司开始发布毫无意义的比较页面时，你就知道他们开始失去计划了。谁在为AI和Loom玩弄Linear、Jira、Rovo？但它仍然没有发布一个单一的比较页面）RT Karri SaarinenYou know a company is starting to lose the plot when they start publishing nonsense comparison pages.Who is juggling Linear, Jira, Rovo for AI, and Loom? 😂(we still haven’t published a single comparison page)
RT ClaudeDevs我们已为所有用户重置了5小时和每周的费率限制。RT ClaudeDevsWe've reset 5-hour and weekly rate limits for all users.
享受更多寓言！ClaudeDevs：我们已为所有用户重置了5小时和每周的费率限制。Enjoy more Fable!ClaudeDevs: We've reset 5-hour and weekly rate limits for all users.
我们什么时候才能停止将自主代理与手写代码进行比较？您看不到与工程师手写汇编相比的编译器。Mehul Mohan：花费了165，000美元以上的Fable 5 API才将Bun从Zig移植到Rust。哇！When do we stop comparing autonomous agents to hand-written code? You don’t see compilers comparing to engineers hand-writing assembly.Mehul Mohan: It took $165,000+ of Fable 5 API to port Bun from Zig to Rust.Wow!
RT Emma我真的很喜欢这件作品。这是一篇罕见的文章，从直观的角度解释了“自我进化的代理产品”在当今的实际工作方式。我的要点是：系统需要通过控制上下文从真实的用户交互中建立有效的数据反馈循环;超越静态反馈RT EmmaI really enjoyed this piece. It’s a rare article that explains, at an intuitive level, how “self-evolving agent products” actually work today.My takeaway is this: a system needs to build an effective data feedback loop from real user interactions by controlling context; move beyond static ben
X是竞技场。一直都是X is the arena. Always has been
人工智能将使所有软件都成为Native。毫不妥协的性能和平台亲和力。Chris Tate：引入Native SDK构建原生应用程序的工具包-热重装-标记+ Zig-即时启动-macOS、Windows和Linux-从头构建的图形处理引擎-内置设计系统和主题-定制组件AI will make all software Native.Uncompromising performance and platform affinity.Chris Tate: Introducing Native SDKThe toolkit for building native apps→ Hot reload→ Markup + Zig→ Instant launch→ macOS, Windows and Linux→ GPU engine built from scratch→ Built-in design system and themes→ Custom compo
更多寓言！ClaudeDevs：我们已为所有用户重置了5小时和每周的费率限制。More Fable!ClaudeDevs: We've reset 5-hour and weekly rate limits for all users.
GPT-5.6现已推出。我们一直在Box AI Complex Work评估中评估模型系列，该评估使用Box AI Agent在各种使用企业文档集的极其困难的任务上测试模型。Sol是GPT-5.5的一大进步，尤其是在需要深入研究的复杂面向数据的任务中 GPT-5.6 is now out. We've been evaluating the model family on the Box AI Complex Work eval, which tests the model with the Box AI Agent on a variety of extremely hard tasks using enterprise document sets. Sol is a big step up from GPT-5.5, especially on complex data-oriented tasks that require deep
RT Box我们在Box的复杂工作Eval上评估了@AIatMeta Superintelligence Lab的Muse Spark 1.1。它与顶级模型具有竞争力，尤其是在结构化数据分析和报告起草方面。要点：模型很重要，但受监管的内容是企业人工智能大规模有用的原因。阅读更多RT BoxWe evaluated Muse Spark 1.1 from @AIatMeta Superintelligence Lab on Box’s Complex Work Eval.It was competitive with top-tier models, especially on structured data analysis and report drafting.The takeaway: models matter, but governed content is what makes enterprise AI useful at scale.Read mor
被放弃的最新人工智能模型正在非常好地处理复杂的知识工作者任务，特别是处理法律、专业服务、医疗保健等复杂的工作领域。Grok 4.5是这里的另一个很棒的条目，尤其是在性价比方面。作为模特gThe latest AI models being dropped are getting insanely good handling complex knowledge worker tasks, and especially dealing with sophisticated domains of work like legal, professional services, healthcare, and more. Grok 4.5 is another great entry here, especially on cost + performance. As models g
RT Sam Lyman新：《纽约时报》报道称，“中国、俄罗斯以及（较小程度上）伊朗试图利用官方媒体将有关美国数据中心的争议变成‘国内破裂点’。'“1月至6月期间，这3个国家的官方媒体提到了RT Sam LymanNEW: The New York Times reports that "China, Russia and, to a lesser extent, Iran have sought to use state media outlets to turn the controversy over data centers in the United States into 'a domestic fracture point.'"Between Jan. and June, state media from these 3 countries mentioned da
🇺🇸斯里拉姆·克里希南：本周，来自@elonmusk / @mntruell的Grok和来自@finkd的Muse Spark之间非常高兴地看到美国的前沿模型生态系统。🇺🇸Sriram Krishnan: between Grok from @elonmusk / @mntruell and Muse Spark from @finkd this week very excited to see an ecosystem of frontier models in America.
RT Ahmad Sadeddin我们今天开源了Sightthound。Sightthound是一个基于Rust的源代码静态漏洞扫描器。它在本地或CI中运行，使用Tree-sitter解析器，并支持基于模式的检测和污染流分析。它还附带了所有规则集，没有付费帐户oRT Ahmad Sadeddin🚀 We open-sourced Sighthound today.Sighthound is a Rust-based static vulnerability scanner for source code. It runs locally or in CI, uses Tree-sitter parsers, and supports pattern-based detection plus taint flow analysis. It also comes with all it's rulesets with no paid account o
X：“法国-摩洛哥比赛后巴黎将陷入火海”赛后巴黎：X: “Paris will be in flames after the France-Morocco match”Paris after the match:
RT Nvidia作为种子轮融资的一部分，我们的融资额超过了1亿美元，并欢迎@nvidia加入我们的资本表。这为我们的发展势头增添了动力：实时STT和TTS，语音翻译，Phonon，所有这些都在几个月内发货，并且在发布之间加快了时间表。它还支持RT GradiumWe’re crossing $100M in funding and welcoming @nvidia to our cap table as part of an extension of our seed round.This adds fuel to our momentum: real-time STT and TTS, Gradium Translate, Phonon, all shipped in a few months and with accelerating timelines between releases.It also supports o
作为一名前电气工程师，在大学里因这样做而陷入麻烦，我非常喜欢这个。Dorsa：工程师在电路板上隐藏了几乎没有人看到的小信息。我开始收集它们并把它们都放在一个地方，享受！As a former electrical engineer who got in trouble for doing this in college, I love this so much ⚡️Dorsa: Engineers hide little messages on circuit boards that almost no one ever sees. I started collecting them and put them all in one place, enjoy!
超级令人印象深刻。Jarred Sumner：在Rusthttps中重写Bun：bun.com/blog/bun-in-rustSuper impressive.Jarred Sumner: Rewriting Bun in Rusthttps://bun.com/blog/bun-in-rust
会在那里进行直播！OpenAI开发人员：OpenAI构建周即将到来。挑战将于7月13日开始。本周，与世界各地的建筑商一起参加现场会议和社区活动。注册是开放的。将想法放在您的积压中。使用Codex构建它。Will do a livestream there!OpenAI Developers: OpenAI Build Week is here.The challenge starts July 13. Across the week, join live sessions and community events with builders around the world.Registration is open.Bring the idea sitting in your backlog.Build it with Codex.
RT安德鲁·安布罗西诺️🐐RT Andrew Ambrosino🐐❤️
所以这意味着.开发人员不工作吗？lolso this implies that...developers don't do work? lol
RT The InformationChatGPT-5.6就像保时捷，而Claude Fable 5就像“曲速驱动”，@Danshipper解释道。Every首席执行官说：“它速度很快，功能强大。它非常便宜……这是你想要的东西的高端。你可能想每天开车。它处理得非常好。感觉很超值。”“WiiRT The InformationChatGPT-5.6 is like a Porsche and Claude Fable 5 is like “warp drive”, @danshipper explains. The Every CEO says: “It's fast, it's powerful. It's very cheap… it's the top end of something that you want. You might want to drive every day. It handles really well. It feels premium.”“Wi
RT claire vo Thomon花了几分钟时间讲述一封写给gpt 5.6 Dan Shipper的现场情书：GPT-5.6 Sun：知识工作的黄金标准https://x.com/i/broadcasts/1qKVmmlvZXWxB📧RT claire vo 🖤on for a few minutes to narrative a live love letter to gpt 5.6Dan Shipper 📧: GPT-5.6 SOL: THE GOLD STANDARD FOR KNOWLEDGE WORK https://x.com/i/broadcasts/1qKVmmlvZXWxB
RT Jonathan Brebner去过@ southpkcomons的人会告诉你它在他们的创始人之旅中有多么有价值。但公司需要一段时间才能让不熟悉SPC的人看得懂。以下是一些更清晰的证据，证明-1到0方法有效：RT Jonathan BrebnerThe people who've been through @southpkcommons will tell you how valuable it was in their founder journeys. But it takes a while for companies to become legible to people who aren't familiar with SPC. Here's some more legible proof that the -1 to 0 approach works:
RT Dheemanthh Reddy有几个时刻你只有回头才能理解。你追踪你现在所处的位置，你就可以确切地看到它的开始。对我来说，这就是SPC创始人奖学金。在奖学金的某个地方，它触动了我：我真正痴迷的是让技术更接近现实RT Dheemanth ReddyThere are a few moments you only understand looking back.You trace where you are now, and you can see exactly where it started.For me, it was the SPC Founder Fellowship.Somewhere in the fellowship, it hit me: what I was actually obsessed with was bringing technology closer to the p