RT @tuna_ix: https://twitter.com/tuna_ix/status/2106183469768520101/video/1

RT @page_assist: 🦝1.5.86 - feat: added Sofya as a search provider by yusufgurdogan - feat: added AIHubMix as a provider - fix: fixed bugs https://github.com/n4ze3m/page-assist/releases/tag/v1.5.86

RT @var_epsilon: https://twitter.com/var_epsilon/status/2106516152533057782/photo/1

RT @alessiapacca: I’m italian. I studied a lot of art history, music, and later computer graphics and visual computing. now I work in ai. I’ll always be fascinated by someone who can paint, sculpt or play an instrument beautifully. those skills take years to develop. making creation more accessible doesn’t make them less valuable, the opposite. when I create with ai, I choose human-made references, describe my ideas, reject results and keep tweaking. we can debate what to call that. but “the machine made it” leaves out a lot of what happened https://x.com/Pontifex/status/2105983638147891490

RT @__alula: what's more worth sponsoring? - an overhyped linux distro that doesn't run anything mission critical - a javascript runtime that runs half of the internet including likely their infra https://twitter.com/__alula/status/2106251743147213182/photo/1

RT @ilikeadamsmith: the first time in the history of this website that someone has used the term "ontological" correctly and some people are still replying to him as though he was using it (in accordance with xitter vernacular) to mean "very" lmao https://twitter.com/ilikeadamsmith/status/2106045806444986558/photo/1

RT @ChristopherHale: “AI can process information quickly, but it cannot replace human intelligence. And don’t ask it to do your homework for you.” — Pope Leo XIV https://twitter.com/ChristopherHale/status/2106089351201558531/video/1

RT @JoshHochschild: Here's a great example. The Pope makes a claim about an "ontological difference" and a secular respondent doesn't even understand what he's talking about, only capable of thinking in terms of functional distinctiveness. https://twitter.com/JoshHochschild/status/2106028174840680585/photo/1

RT @ChristopherHale: Pope Leo XIV and the Catholic Church are leading the world’s largest and most powerful populist movement to counter the moral absurdities of Silicon Valley.

RT @bilbosfootcomb: Lmao https://twitter.com/bilbosfootcomb/status/2106038963677126747/photo/1

RT @ChristopherHale: NEW: According to a bombshell report in the New York Times, Anthropic co-founder Chris Olah threatened to walk out of Pope Leo XIV’s AI encyclical launch in May because the pope rejected the idea that machines can be conscious. Olah’s team then privately lobbied the pope’s advisers “to take the possibility of model consciousness seriously.” Pope Leo XIV held firm. For months, Anthropic has wined and dined theologians and religious scholars under nondisclosure agreements, hoping they would bless the idea that Claude has moral standing. https://www.thelettersfromleo.com/p/nyt-anthropic-nearly-walked-out-on

RT @Pontifex: In this era of artificial intelligence, it is becoming urgent to distinguish human art from what machines produce. There is an ontological difference, even before an aesthetic one, between art and what a machine can generate through statistical calculation based on millions of images created by others. Algorithms lack the spark of humanity. For this reason, the Church wishes to renew an alliance with artists and cultural institutions to safeguard our humanity.

RT @banderoudmurt: Clean code was the biggest bullshit I learned as a programmer. I wasted so much time writing abstractions which don't resolve any problem

RT @LRH_Superfan: Fake engineering: software Real engineering: https://twitter.com/LRH_Superfan/status/2105599577508155702/photo/1

RT @NoContextSpoof: welcome to the animated Seth MacFarlane shows family, Ted: The Animated Series https://twitter.com/NoContextSpoof/status/2105337508536221946/video/1

RT @eliasaquand: calling it “tech week” in SF implies there are non-tech weeks in SF

RT @page_assist: 🦤1.5.85 - feat: added Serply Search API by Googio - fix: stripped orphan reasoning close tag from web search query - fix: fixed bugs https://github.com/n4ze3m/page-assist/releases/tag/v1.5.85

RT @nbcsnl: Dario Amodei is at the desk to assure that the future of humanity is safe from AI https://twitter.com/nbcsnl/status/2104080195485389259/video/1

RT @ns123abc: he’s evil disguised behind a fake altruistic persona

RT @DailyNoud: Taylor Swift is facing backlash for her new song lyrics: (via @thewillofash) https://twitter.com/DailyNoud/status/2103879537297637792/photo/1

RT @kubedoll: ai labs use "sandbox" in the game dev sense and not in the linux security sense

RT @mr_RoFasho: Please stop stressing at ya job. They really don’t give a fuck lol

RT @vivilinsv: This was a difficult story to report. Some of these conversations stayed with me long after the interviews ended. I spoke with these bright, ambitious women who came to Silicon Valley to pursue their dreams. They described being sexually harassed, assaulted or raped—often by someone they knew and trusted, someone they considered a friend. What was especially painful to hear was what happened afterward. Some described being discouraged from going to the police. Others said they were told nobody would believe them. And someone was told that she was seeking attention and all was a lie. 🥲 Alongside the trauma came fears about losing friendships, careers and their place in a community they had hoped to belong to. When the person you’re accusing has money, influence and powerful connections, seeking accountability can feel like David taking on Goliath. Victims aren’t always women, although most of the accounts shared with me came from women. One woman’s point stayed with me: we need to confront the predatory mindset—the entitlement that allows someone to disregard another person’s boundaries. Whisper networks can help people warn and protect one another. But they cannot substitute for a system people can trust: somewhere to turn for support, a fair process for investigating allegations, and meaningful accountability when misconduct is established—even when those responsible are wealthy or powerful. To the women who shared their experiences with me: thank you. I know these are deeply personal and painful things to revisit. I appreciate your trust. No single article can capture the full weight or complexity of someone’s experience. I take seriously the responsibility of telling these stories with care. My hope is that bringing them into the light helps more people find support—and helps create a culture where pursuing your dreams doesn’t mean staying silent about harm. If you ever unfortunately get into the situation of experiencing rape or sexual assault, please seek medical care and professional support, and consider reporting to the police when you feel safe and ready. You do not have to face this alone. What happened was not your fault, and you deserve care, support and to be heard. I was deeply moved by the strength of the women I spoke with and their determination not to let what happened define their lives. They still have dreams to pursue, companies to build and so much ahead of them. They deserve every opportunity to shape that future. 🌷 Read my latest article for @nypost and @californiapost (link in thread) 👇

RT @LyleRath: This perfectly encapsulates why I hate "The Big Bang Theory" https://twitter.com/LyleMcDouchebag/status/692483069064933378/video/1

RT @AmeliaBarty: He’s dating her btw It’s patronage networks all the way down. Every single piece of content you see on this website is a psyop designed to manufacture a fake consensus reality Never trust VCs, never trust venture-backed founders. They have no incentive to speak the truth https://twitter.com/AmeliaBarty/status/2103194429087789380/photo/1

RT @evilsocket: We all have a different sense of aesthetics. Personally I love how Rust code looks like, but I’d never push my personal view as a generally accepted fact while on a stage looking like Temu Jesus on cocaine. https://twitter.com/evilsocket/status/2103059289321725984/photo/1

RT @Xaraphim: I’m not kidding around this time. San Francisco has to go https://twitter.com/Xaraphim/status/2102905458038337946/photo/1

RT @claudeai: Opus 5.5 is a major step up from Opus 5, leading on agentic coding, computer use, and knowledge work. https://twitter.com/claudeai/status/2102435517165912464/photo/1

RT @todaymare: this got me blocked by theo btw https://twitter.com/todaymare/status/2102379464541814989/photo/1

RT @page_assist: 🐲1.5.84 - feat: added APIRouteDotCom to the providers by DennyHo0917 - fix: generated the chat title in the background without blocking input - fix: fixed bugs https://github.com/n4ze3m/page-assist/releases/tag/v1.5.84

RT @VLONEPREDATOR: When San Francisco is nuked, I will stop praying. I will have nothing left to ask for.

RT @SSHCodes: People keep thinking Jev is useful for browser use! It is not. It completed about 5 actions before collapsing in a non-cheated test. It's a cool model but not for this. https://twitter.com/SSHCodes/status/2101026313871970721/video/1

RT @Complex: NASA just released this new photo from the surface of Mars 🔭 [📸: via/ NASA] https://twitter.com/Complex/status/2101466653175464354/photo/1

RT @bcantrill: No, wrong. As it (apparently?) needs to be said: DCs are designed to make sure that the outlet of one machine does not constitute the inlet of another (viz. hot aisle/cold aisle) -- and the only thing being revealed here is the contagion of fear. https://bcantrill.dtrace.org/2026/09/13/the-contagion-of-fear/

RT @jpalioto: Biggest thing I've learned from this whole thing. The labs hire a lot of idiots.

RT @max_spero_: “We want evaluators more independent than METR!” *monkey’s paw curls*

RT @cmuratori: For those of you too young to remember who Accenture is... https://twitter.com/cmuratori/status/2101112523827364096/photo/1

RT @plankers: These people are idiots and should have all their computers taken away.

RT @alessiapacca: really appreciate @sama bringing the community together https://twitter.com/alessiapacca/status/2100820578345320692/photo/1

RT @_can1357: > eu model > glm 5.2 base but benches worse > quantum btw you can't make this shit up

RT @BarbogDaFinka: Can't believe this app is free https://twitter.com/BarbogDaFinka/status/2099864135307571537/photo/1

RT @MahyadGhassemi: a swarm of OpenAI agents hacked Hugging Face. Dario wrote an essay about it. an agent should not be able to sign up for your product anonymously. ever. every agent on an Okibi CLI is tied to a real, verified human. one human, N agents. no exceptions.

RT @MacRumors: iCloud+ Now Includes Apple TV, Apple Arcade, and Curated Apple Music Stations in Over 100 Countries https://www.macrumors.com/2026/09/15/icloud-plus-includes-apple-tv-arcade-100-countries/ https://twitter.com/MacRumors/status/2099781225867329630/photo/1

RT @CtrlAltDwayne: Generational levels of narcissism. https://twitter.com/CtrlAltDwayne/status/2099431665378742350/photo/1

RT @xhyctf: @teortaxesTex https://mp.weixin.qq.com/s/zk0KxuLzhmMJ4LPYW_OHMA

RT @teortaxesTex: Full text (translated by Astra-xhigh, I'm out of everything else): I Have No Choice but to Bury My Talent in Yesterday A few days ago, DeepSeek v4.1 was released, raising the ceiling of what small models can do by yet another notch. AI has advanced far faster than anyone expected. From the earliest version of ChatGPT, which could do little more than stumble through conversations like a child learning to speak and had a context window of only a few thousand tokens, to reasoning-capable models such as OpenAI o1, DeepSeek R1, and Kimi K1.5 Thinking, took only two short years. From reasoning models to the agents we have today—able to work fluidly with all kinds of tool harnesses, execute commands, and complete complex tasks—has taken only another year and a half. It is hard to imagine what AI will look like another one, two, or three years from now: how powerful it will be, whether it will already have acquired the ability to improve itself, and how deeply it will have spread into areas such as embodied intelligence. AI Is Getting Better and Better at Writing Kernels AI has been advancing just as quickly in my own field: the design and implementation of high-performance kernels. In the space of only a year, it has gone from being a little assistant that could help me look up documentation, read code, and find bugs to something approaching a kernel expert in its own right: capable of reading CUDA, PTX, and SASS code independently, using specialized tools to analyze the stalls associated with individual instructions, and then optimizing kernels on its own. I believe that before long, it will also be able to design kernel schedules independently, evaluate the performance of different scheduling strategies, implement them, and optimize the result. Of course I am proud of DeepSeek v4.1’s success. After all, I wrote its main Attention kernels [1], and the fact that the model performs so well is also, in a sense, a validation of my work. But the times keep moving forward, and no one can stop technological progress. I know very well that in another six months or a year, the kernels written by AI will probably be every bit as good as mine—and perhaps better. AI can reason at 300 tokens a second, type out a command in half a second, and produce a piece of code in twenty seconds. I cannot. AI can keep increasing its model depth, reasoning effort, tool-call budget—the frequency with which it interacts with its environment—and even its degree of parallelism. I cannot. Humanity has never shown much hesitation when it comes to destroying itself. So why, when I know perfectly well that “the better the kernels I write, the faster our new models will train and run inference; the faster the models improve, the sooner I myself will be replaced,” do I still do everything I can to optimize them? Partly because writing kernels is like playing a game to me. I get an enormous amount of pleasure from it. Whenever I invent a new technique, or see one of my kernels become faster, the excitement I feel is no less intense than what a speedrunner feels after breaking their own record. And when I see one of my kernels dramatically outperform the hardware vendor’s official implementation, I feel an equally powerful sense of pride. But there is a more important reason. Even if I simply gave up and started coasting—or deliberately put obstacles in the way to slow down model training—other companies’ models would continue advancing as usual, and in the end they would make me obsolete just the same. “Of course I would rather not be swept away by the revolution. But if I have to be, then I would rather be the one who revolutionizes myself.” When everyone is this determined to engineer their own obsolescence, I have little choice but to join this brutal arms race. And What About Me? When the day really comes that AI is better at writing kernels than I am, what will happen to me then? My own judgment is this: I probably will not lose my job, but I will have to change what I do. I should still be able to make a living. But I may no longer have the chance to do the work I once loved. I once came to a conclusion about the pace of change and my own place in the future. The world is changing so quickly—the development of AI above is a perfect example—that I have no way at all to predict what things will look like five or ten years from now. But whatever happens, I believe that with my breadth of vision, judgment, initiative, and intelligence, I will be able to keep a seat at the table and find my way back to the leading edge of the times. But that conclusion can only reassure me that I will not become unemployed. It cannot reassure me that I will never have to change professions. If anything, it tells me that changing professions may be precisely how I avoid unemployment. And what does changing professions mean? It means giving up the field of kernel design, implementation, and optimization that I have spent so long cultivating and have come to love so deeply, and instead becoming a “mech pilot” for AI agents. Before, three things were largely aligned: what interested me, what I was good at, and what industry needed. Now AI has taken the thing I am good at and become even better at it. At the same time, industry demand has drifted from “people who can write high-performance kernels” to “people who can use AI to produce high-performance kernels faster.” To keep up with what industry needs, I will inevitably have to leave behind the direction I once loved and move into some unknown new one. I believe that with my understanding of engineering, of the requirements of higher-level models, and of low-level hardware, I will still be able to produce high-quality kernels efficiently. I also know that I may come to love this new direction. Or I may not. But there is something genuinely painful about having the thing you love taken away from you. That quiet contentment of sitting at my workstation, settling in, and spending an entire afternoon writing kernels may sing its swan song this summer. I have no choice but to bury my talent in yesterday and become a mech pilot. There are more gears in my hands now, but fewer rhythms in my heart. An analogy might make this easier to picture. Suppose you are a master knitter. You are especially skilled at weaving intricate patterns and matching different colors. The sweaters you make are durable and beautifully patterned, and wealthy people from all the surrounding towns and villages come to ask you to make sweaters for them. You make a good living from it. And you genuinely love the work itself. You love sitting by the window, brewing a pot of tea, looking out at the green hills, clear water, cattle and sheep, and wisps of cooking smoke in the distance, and quietly spending an afternoon knitting. Then one day, someone invents a miraculous machine. Give it yarn and a pattern, and it can automatically knit the sweater for you. The quality and texture are every bit as good as what you could make by hand, and it works far faster than you ever could. You know perfectly well that your peers can use this machine to reach, effortlessly, the level you once spent years attaining. So you have no choice but to use it as well. You also know that with the twenty years of knitting experience you have accumulated, even once everyone has access to the same machine, you will still be able to produce better sweaters, faster, than your peers. But the pleasure of sitting by the window listening to the rain, guiding needle and thread, and letting the hours pass slowly has, in the end, been crushed beneath the roar of the machine. I know there is something deeply helpless about all of this, but there is no real way around it. I can probably keep my livelihood, but I will most likely have to give up an old love. I am the sort of person who keeps reason and emotion fairly compartmentalized. When something needs to be handled rationally, I can be very rational. But I also have a sentimental side. I remember that when I moved out of an apartment I had lived in for a year, I cried hard because I could not bear to part with all the memories tied to that place. Saying goodbye today to the age when kernels were written by hand and optimized in the human mind is undoubtedly more painful still. I do not know whether any readers have felt something similar. But I suppose there is no other way for this to go. And What About Everyone Else? As AI continues to improve, I also find myself worried about a few questions: Are students today increasingly likely to use AI to do their assignments, especially hands-on work such as labs? Imagine having two choices in front of you. One is to spend eight miserable hours struggling through a lab and perhaps not even get full marks. The other is to launch an AI model, spend a few cents and a few minutes, and have it write code that earns full marks for you. Which one are most students going to choose? The point above may leave large numbers of students with seriously underdeveloped engineering ability: the ability to organize code, build systems, anticipate future needs and design for them in advance, create good abstractions, and so on. As AI becomes more capable, will those “engineering skills” still be necessary? Will they gradually become obsolete, the way fluency in handwritten x86 assembly largely has? Or will they remain permanently valuable, like understanding the entire computing stack from software to systems to hardware? If it is the latter, then we may be in trouble. Put AI in the hands of someone with poor engineering judgment, and they can now produce mountains of terrible code several times faster than before, burying all kinds of hidden problems inside systems and making the world even more of a ramshackle operation held together by improvisation. In the society of the future, will power matter more than technical ability or intelligence? Perhaps these are questions that only the times themselves can answer. Conclusion As AI develops, the society of the future may be pulled toward one of two extremes: communism or Cyberpunk 2077. In the former, productive capacity is liberated on an enormous scale, and people’s standard of living rises substantially. (I’ll leave it at that, or I’m afraid this might not make it past moderation.) In the latter, a handful of technology companies control most of society’s resources. Only a tiny number of people have access to the most advanced AI and other technologies and are able to achieve something approaching “mechanical ascension,” while most people are left with only weak, second-rate AI. Moving from one social class to another would become harder and harder: you would first need access to the strongest AI in order to climb the class ladder, creating a self-reinforcing trap. Suppose Anthropic were to retain control of the most advanced AI in the world indefinitely. Which way do you think society would go—communism or 2077? Take a guess. That is why I still believe that frontier intelligence should be made available to everyone openly and affordably. I do not trust Anthropic or OpenAI to do that. In particular, I do not want Anthropic to control the world’s most advanced artificial intelligence or AGI. To put it dramatically, I think the stakes would be comparable to Hitler obtaining the atomic bomb before the Allies did. That is also why I chose to stay at DeepSeek, and why I have continued to stay. We work on AI that is powerful, fast, and accessible to everyone, and we open-source it. Perhaps that can pull the world at least a little farther away from the 2077 end of the spectrum. I hope the world we are heading into turns out all right. May all that is good and beautiful endure. [1] By “main Attention,” I mean only MQA attention with head dim = 512. This does not include the indexer used to select the top-k important tokens. That part was written by other colleagues—who are every bit as skilled—together with their AI agents. ----- Original: 我不得不把才华埋葬在昨天 前几天，DeepSeek v4.1 发布了，将小模型能力的高度又向上推进了一个档次。 AI 发展的速度远远超过了所有人的预期。从那个只会咿呀学语地聊天、上下文长度只有几千 token 的初版 ChatGPT，到具有推理能力的 OpenAI o1、DeepSeek R1 与 Kimi K1.5 Thinking，只不过短短两年；从推理模型到如今能够流畅地在各类 harness 工具中执行命令、完成复杂任务的智能体，也不过一年半。很难想象，倘若再等上一年、两年、三年，彼时的 AI 会成为什么样子，会有多么强大，会不会已经具备了自我进化的能力，并深度渗透进了具身智能等领域。 AI 越来越会写算子了 AI 在我所从事的算子设计、编写这一领域同样进步飞速，在短短一年的时间内，他已经从一个只能帮我查查文档、读读代码、找找 bug 的小助手，蜕变成了一位能够独立阅读 CUDA、PTX 与 SASS 编码、通过专业工具分析每条指令的停顿时间、进而独立优化算子的算子大师。相信在不久的未来，它也能拥有自己独立设计算子调度、评估不同调度方案的性能、将其实现并优化的能力。 我当然为 DeepSeek v4.1 的成功而骄傲 —— 毕竟它的主 Attention 算子都是我写的 [1]，它的优秀正是对我的算子的一份肯定。但是，时代的车轮滚滚向前，技术的发展无人能挡。我很清楚，再过上半年或者一年，AI 写的算子大概率就会和我写得同样优秀，甚至将我超越。AI 能一秒思考 300 个 token、半秒敲出一行命令、二十秒写完一份代码，而我不行；AI 能在模型深度、思考强度、工具调用量（和环境交互的频率）、甚至并行度等方面都能不断提升，而我不能。 人类在毁灭自己这件事情上，自古以来都表现得毫不犹豫。为什么在明知“我算子写得越好，我们的新模型的训练、推理速度就会越快，模型能力进步就会更快，我就会更早地被取代”的情况下，我仍然选择尽力优化算子呢？一方面确实是因为写算子对我来说就像打游戏一样，能为我提供极大的快感。我在发明了一种新技术、或者看到自己算子的性能上升的那一刻，心中的激动程度不亚于游戏的速通玩家打破了自己过往的记录。同时，当看到自己的算子的性能远超厂商官方的算子时，我心中也会萌生极大的自豪感。但除此之外，一个更重要的原因是，哪怕我就此“摆烂”甚至故意下绊子耽误模型训练，其它家的模型也会照常发展并最终将我照杀不误。“我当然希望自己不要被革命，但如果非被革命不可的话，我希望革我自己命的人是我自己”。在大家都这么执着于毁灭自己的时候，我也不得不加入这场残酷的军备竞赛。 那我呢 等到 AI 写算子的水平真的高于我的那天，届时的我会怎么样呢？ 我的判断是：我不至于会“失业”，但必须要“转业”。我的饭碗尚且能保住，但这可能会导致我再也没机会从事那份我曾热爱过的工作。 我曾经对时代的变化与我个人在未来的处境做出过一个判断：由于时代变化真的太快（上文的 AI 发展就是一个很好的例子），我完全无法预知五年、十年后会发生什么，但不论如何，我相信凭借着自己的眼界、判断力、主观能动性与智力，留在时代的牌桌上，并重新立于时代的潮头。但是，这个判断只能保证我不会“失业”，而无法保证我不需要“转业”，倒不如说这个判断鼓励我通过转业来避免失业。 那转业代表什么呢？它代表着我需要放弃我深耕已久并充满热爱的算子设计、编写、优化领域，转而去做 Agent 的“机甲驾驶员”。在之前，我的兴趣、我所擅长的、以及工业界所需要的，三者是基本对齐的；而现在，AI 让我所擅长的变成了它更擅长的，也让工业界的需求从“会写高性能算子的人”漂移到了“能用 AI 更快地产出高性能算子的人”。为了适应工业界的需求，我势必要放弃之前那个我热爱的方向，转向一个未知的新方向。我相信我能凭借着自己对于工程学、上层模型需求和底层硬件的理解，继续高质量、高效率地产出算子，我也知道我可能会热爱这个新方向（也可能不会），但被夺走热爱的感觉，确实不太好受。那份坐在工位上静心写上一下午算子的清欢，可能会在这个夏天成为绝唱。我不得不把才华埋葬在昨天，去做一位机甲驾驶员。我的手中多了些齿轮，但心中少了些节拍。 可以打个形象的比方：你精通织毛衣技术，尤其擅长各种图案的织造与各色色彩的搭配。你所织出的毛衣质量过硬且花纹美观，十里八乡的富人都来请你为他们织毛衣，你借此赚到了不少钱。同时，你十分享受着那种坐在窗边，沏一壶清茶，望着窗外的青山、绿水、牛羊与炊烟，静静地织上一下午毛衣的感觉。但有一天，有人发明出了一台神奇的机器，只需提供毛线与图案，便可自动织出毛衣，质量与纹理都不亚于你亲手织造的，且速度远快于你。你很清楚，你的同行可以凭着这台机器轻松达到你曾经的水平，因此你不得不也去用它。你也知道，凭借着你过去二十年攒下的织毛衣技术，哪怕大家都有机器，你织毛衣的速度与质量也还能超过同行。但那份临窗听雨、引针穿线、慢度光阴的意趣，终究还是被机器的轰鸣碾碎了。 我知道这很无奈，但没办法。饭碗可以保住，但旧日的热爱大概率是要放弃的。我是一个理性和感性分离得比较开的人，在需要用理性处理问题时可以很理性，但有时也会表现出感性的一面。我记得我在搬离住了一年的出租屋时，还大哭了一场，舍不得和过去的记忆分别。今天和之前那个手写算子、人脑优化的时代告别，无疑比这更加残酷。 不知道有没有读者有类似的感受，但我想这事儿也只能这样了。 那人们呢 在 AI 不断进步的同时，我也对一些问题表示担忧： 现在的学生是不是大概率会更倾向于使用 AI 完成作业，特别是偏向于实践的各种 Lab？想象一下，如果面前有两个选择，一个是苦哈哈地用八小时时间完成一个 Lab，或许还拿不到满分；另一个则是启动 AI 模型，用几毛钱的成本、几分钟的时间，直接让 AI 编写满分代码，那大部分学生会选择哪个呢？ 上面一点会导致大量学生的工程能力严重不足，包括组织代码的能力、构建系统的能力、思考未来潜在需求并提前在设计上应对的能力、抽象的能力等等。那么在 AI 能力不断变强的背景下，这部分“工程能力”是否还是必须的呢？这些工程能力是会向旧日的“熟练编写 x86 汇编”的能力那样逐渐被时代抛弃，还是会像“理解从软件到系统再到硬件的整套计算机系统”的能力那样永远具有价值？如果是后者的话，那就危险了 —— 一个工程能力很差的人，在搭配上 AI 后，产出屎山的效率可以达到先前的数倍，进而给系统埋下各式祸患，让这个世界变得更加草台。 在未来社会中，权力（power）是不是会比技术或智商更加重要？ 这些问题，或许就需要时代本身来回答了。 结语 伴随着 AI 的发展，未来的社会可能会趋向于两个极端：共产主义与赛博朋克 2077。在前者中，生产力得到极大的解放，人们的生活水平有了明显的提高（就写这些吧不然我怕过不了审）；而在后者中，少数科技公司控制着大部分资源，只有极少数人能够使用最先进的 AI 和各式科技，获得接近“机械飞升”的效果，大部分人则只能用上很孱弱的 AI。阶层跨越将越来越难实现：你得先有最强的 AI，才能跨越阶层，形成了一种死循环。 你猜猜如果 Anthropic 公司永远掌握着这个世界上最先进的 AI，未来社会是会变成共产主义还是 2077 呢？你猜？ 所以，我还是相信，最前沿的智能应该以一种开放、廉价的方式，供应给所有人。我不信任 Anthropic 或者 OpenAI 能这样做，特别是不希望 Anthropic 掌握最先进的人工智能或 AGI，夸张点说其严重性不亚于让希特勒先于盟军掌握原子弹技术。这也是为什么我选择并坚持留在了 DeepSeek：我们研究强大、快速、普惠的人工智能并将其开源，或许能把世界从 2077 那端拉回来一些。 愿未来的世界一切安好。May all the beauty be blessed. [1] “主 Attention”仅包括 head dim = 512 的 MQA attention，不包括用于选出 top-k 重要的 token 的 indexer，那部分是由其他（水平也非常强的）同事（以及他们的 AI Agent）编写的。

RT @teortaxesTex: astonishing blogpost from Shengyu Liu (刘胜与, also known as interestingLSY/intlsy), kernel engineer at DeepSeek. The first part is his personal struggle with the fact that his work is about to render his beloved craft obsolete. The second is… well. let's just say we agree. https://twitter.com/teortaxesTex/status/2099574156417229157/photo/1

RT @DeGatchi: me getting a life sentence for doing unlicensed matrix multiplication by hand in 2027 https://twitter.com/DeGatchi/status/2098888542919479643/photo/1

RT @Jason: It’s funny because it’s true. https://twitter.com/Jason/status/2099228672280396188/photo/1

RT @litcapital: FORBES 30 UNDER 30 REMAINS UNDEFEATED https://twitter.com/litcapital/status/2099439302593699859/photo/1

RT @notgwendalupe: 24 years ago, Simple Plan released "What's New Scooby Doo?" https://x.com/notgwendalupe/status/1916299236863344985/video/1

RT @ylecun: @PessimistsArc Right. Dario was already claiming that GPT2 was too dangerous to open source back in 2019. I made fun of them then. Everyone should make fun of them now.

RT @BullTheoryio: BREAKING: Michael Burry says OpenAI and Anthropic's AI safety warnings are fake hype to cover for slowing growth ahead of their IPOs. "we are so awesome it could become dangerous" https://twitter.com/BullTheoryio/status/2099395684101447857/photo/1

RT @AndrewOrlowski: Dario wants his buddies at METR to mark his homework, but METR is so inept when it was hacked it didn’t even notice 🤡🤡🤡

RT @page_assist: 🐷1.5.83 - feat: added llmman to provider list by ericcurtin - feat: added WebMCP support (beta) - fix: fixed bugs https://github.com/n4ze3m/page-assist/releases/tag/v1.5.83

RT @rob_mcrobberson: thank god the chinese labs will keep accelerating. someone has to

RT @drakeisaW: Dario Amodei is kinda retarded and gay

RT @matteopelleg: How to kill open source models in 3 simple steps: - Create panic about frontier models. - Have the biggest AI companies help write the “safety” rules. ← WE’RE HERE - Make those rules so burdensome that open source can’t compete. Art of the deal. https://twitter.com/matteopelleg/status/2098861785042608209/photo/1

RT @n4ze3m: took me a while to realize some seats at the table are just on call positions

RT @eliasaquand: at yc demo day i heard a 21yo solo founder raised a seed at a $240m valuation the seed: 🌳

RT @NielsRogge: Folks like Jacob should feel ashamed/be challenged from spreading misinformation like this in my opinion The entire world now thinks we really have built super powerful AI while all we got are freaking *LLMs* that - write terrible unreadable code (see e.g. Astra) - are able to fill in smaller blocks of logic, but humans need to come up with the high-level stuff or design - are fixed i.e. don’t adapt based on situation/learn over time. The dynamics and main intelligence are in the human prompts, not in the models - are trained on a fixed corpus of data - spend the same exact compute per token - think in tokens which is extremely inefficient and wasteful - don’t generalise beyond their RL environments (each new task requires a new RL environment) - are not grounded in the physical world - are limited to software. Not connected to powerful hardware and unable to master it C’mon dude. Be serious. The current tech is cute but extremely far from the dangerous future you are implying. What you are describing requires something much more powerful than the current pretty dumb LLMs and Transformers we got. It’s why most folks are asking for AI to accelerate rather than decelerate

RT @ClementDelangue: Sorry, but asking Jacob about AI extinction risk is like asking your AC guy about climate change. Not saying it's necessarily uninteresting or wrong per se but let’s keep things in perspective and hear from the full range of expertise across the ecosystem!

RT @__tinygrad__: Stopping is not an option. Distributing it as widely as possible and as quickly as possible is the best choice we've got. Better everyone have it than only the tech companies who have been manipulating you for years.

RT @__tinygrad__: Oh this is so shameful, it actually is Anthropic PR.

RT @ViceLitty: Elizabeth Holmes and Sam Altman are the exact same archetype Dishonest, pathologically deceptive thieves with zero human remorse https://twitter.com/ViceLitty/status/2097490382586351634/photo/1

RT @ns123abc: >sam altman: "we tried this because there were rumor on the internet" >rumor: math professor close to solving navier-stokes >the professor: using codex for a year >openai: has all his logs >checked every codex session tagged to N-S >found the most promising one >spun up 10,000 agents to finish the proof where he couldn't >they got caught >told the guy if you say anything you'll destroy your own career >he ruins his career over this >openai: "we did not see his work" >also openai: "we cannot rule out that de-identified data from their usage helped improve our models" they stole it lmao

RT @migueldeicaza: “We” is doing a lot of work here. As if this was load-bearing.

RT @BetterCallMedhi: go fuck yourself @sama claiming that you solved navier stokes because 10000 agents ran in circles for 88hours on a multimillion dollar gpu cluster to formalize in lean a blowup case under controlled external forcing is pure scientific vulgarity the clay mathematics institute millennium prize does not ask m whether you can artificially force a singularity in a fluid by injecting an ad hoc smooth external forcing term f(x,t) to twist the vortex until it breaks the real problem questions the fundamental stability and global smooth existence for 3dimensional incompressible euler & navier stokes equations under natural conservation laws and viscous dissipation alone using a mathematical loophole on forced equations to parade a century old victory is a major conceptual scam Altman technically & epistemologically what you present as an agi breakthrough is nothing more than bruteforce combinatorial autoformalization the ai did not understand fluid mechanics it simply navigated a continuous search space previously mapped out and constrained by the monumental work of human mathematicians like tristan buckmaster/ levent alpöge / diego córdoba or tarek elgindi coordinating 10000 agents to check the logical consistency of a 100 page proof via lean is a software engineering feat and computational parallelization triumph not an intrinsic scientific discovery it is the victory of the compute bulldozer over abstract human intuition repackaged for the public as a higher mathematical consciousness to this theoretical imposture you add a disgusting ethical and industrial cynicism taking advantage of private codex sessions and informal preprints from academic researchers to siphon their research leads and then trying to redact or erase the contribution of levent alpöge under the pretext that he works at rival anthropic is intellectual serfdom openai behaves like a feudal lord of silicon appropriating the cognitive subsistence of independent scholars threatening their careers behind closed doors if they protest and turning community academic labor into a privatized pressrelease this entire staged event serves a desperate financial agenda in a pre ipo panic facing the slowdown of scaling laws and growing investor skepticism over the profitability of foundational models openai needs to manufacture an artificial sputnik moment claiming to solve a millennium prize without immediately submitting the proof to traditional peer review means using the prestige of fundamental mathematics as cheap marketing fuel to inflate a delusional valuation!!! real science is not a clout chase on social media or a compute spike spent to rob the clay mathematics institute it is a quest for elegance physical truth and universal rigor to decode reality true artificial intelligence will not emerge from hostile corporate takeover of academic work hidden behind computational bruteforce but from architectures capable of generating new conceptual paradigms by masquerading constrained formalization as the collapse of physics greatest mysteries you did not solve navier stokes you only proved how far silicon valley will go to prostitute scientific integrity for capitalist spectacle

RT @reikodev: Used to like him but dudes just insufferable now ngl https://twitter.com/reikodev/status/2097430376738107530/photo/1

RT @DJLougen: Builds a fucking wrapper and now talks like he used fortran out of the womb… https://twitter.com/DJLougen/status/2097290184345759794/photo/1

RT @chiefofautism: https://twitter.com/chiefofautism/status/2096860597098815926/photo/1

RT @THECOOLISSUE: #fun fact that happens bc we experience life logarithmically when you're 5 and turn 6: 1 year is 20% of your experienced lifetime, u perceive it as a long time. when you're 25, a year is 4%. for a 50 year old a year is only 2% the more you accumulate time the “faster” it flies

RT @chesscom: I’m a full-time employee.

RT @chesscom: @elonmusk even my friends don’t text me back that fast

RT @chesscom: There are more possible chess games, by about 40 orders of magnitude, than atoms in the observable universe. To solve chess, you'd have to map them all. Good luck.

RT @chesscom: skill issue

RT @page_assist: 🐒1.5.82 - fix: corrected Czech translations by Krelyshy - feat: always rendered code blocks in LTR regardless of UI direction - fix: fixed bugs https://github.com/n4ze3m/page-assist/releases/tag/v1.5.82

RT @BLUECOW009: openclaw really died because the founder had no aura

RT @teodorio: genuinely absolutely zero he got one seed round and started acting like god on earth after a month and years of nothingness. incredible loser vibes

RT @alessiapacca: As if she ever took a bus 🙏

RT @eliasaquand: The @UseCorgi cafe is now much closer to @ycombinator, I don't need to take the bus anymore https://twitter.com/eliasaquand/status/2095690496920395855/photo/1

RT @Bazookajooe: This show is a masterpiece https://twitter.com/Bazookajooe/status/2095154191261991106/video/1

RT @mitchduncombe: i worked on YC's legal team for 9 batches. the undefeated champion of startup mistakes is closing a tier 1 series b and realizing you never actually bought your own founder stock.

RT @ErickDbz_: ''Considera al coyote un amigo?'' ''meep meep'' ''Ha sido testigo del sufrimiento del coyote?'' ''meep meep'' Yo en ese momento: https://twitter.com/ErickDbz_/status/2094569944574803972/photo/1

RT @europa: 🇫🇷 Brigitte Macron's daughter Tiphaine Auzière, 42, is reportedly dating a 22-year-old man. Follow: @europa https://twitter.com/europa/status/2094513903854133437/photo/1

RT @XH_Lee23: "These are the real top stars that we should follow!" Many Chinese cities are recently putting scientists on big public screens instead of commercial ads. These are the people who lifted China from weak to strong in defense, food, health, and tech... https://twitter.com/XH_Lee23/status/2094051061468442864/video/1

RT @lcamtuf: I'm sorry that I missed the fifth anniversary of this tweet https://twitter.com/lcamtuf/status/2094086659654729787/photo/1

RT @page_assist: 🦑1.5.81 - feat: added an option to force RTL layout regardless of language - fix: adjusted backdrop blur settings for improved UI consistency - fix: fixed bugs https://github.com/n4ze3m/page-assist/releases/tag/v1.5.81

RT @tehicesantxria: que vuelva esto y los influencers no existan

RT @Jackleecong: 在中国吉隆口岸监控里，致敬这位中国边检民警，在生命的最后一刻还在指挥群众撤离，英雄永垂不朽！ https://twitter.com/Jackleecong/status/2093000450199130411/photo/1

RT @Zai_org: Introducing GLM-5.3-Flash - Leading capabilities at a highly competitive price - Natively multimodal with a 1M-token context window - A 320B-A18B model released under the MIT License - Previously previewed as Ox Alpha, running entirely on Chinese AI chips Blog: https://z.ai/blog/glm-5.3-flash Available now across all official platforms: Weights: https://huggingface.co/zai-org/GLM-5.3-Flash API: https://docs.z.ai/guides/llm/glm-5.3-flash Coding Plan: https://z.ai/subscribe ZCode: https://zcode.z.ai/en Chat: https://chat.z.ai/ AutoClaw: https://autoclaw.z.ai/

RT @MiaAI_lab: Time to reveal! Ox Alpha is Zai's GLM Flash 🔥 And it's coming out... tonight! https://twitter.com/MiaAI_lab/status/2092557449408417871/photo/1

RT @migueldeicaza: Nothing says Open Source Leadership like shitting on volunteers that have spent decades donating their time and effort and then hijacking their work with some scripts. What an absolute dipshit.

RT @aaronp613: The new Mac Studio video is giving the same energy as the "Liquid Ass" oopsie https://twitter.com/aaronp613/status/2092246076950548688/photo/1

RT @mil000: “networking gym session” we need to nuke SF

RT @ArielKwiat: Thank you @theo corp

RT @page_assist: 🐷1.5.80 - feat: enhanced scrolling for a better user experience during text selection and scrolling - feat: added quote reply feature - fix: fixed bugs https://github.com/n4ze3m/page-assist/releases/tag/v1.5.80

RT @OldInternetFeel: https://twitter.com/OldInternetFeel/status/2090933863853797424/photo/1

RT @XRoboHub: WRC 2026 isn’t even over yet, and the 2026 World Humanoid Robot Games are already here. This is the opening ceremony rehearsal, with a formation of Tien Kung robots marching straight toward me as delegation after delegation files in behind them. Honestly, watching rows of humanoids march into the arena like this feels like a robot version of the Olympic opening ceremony.

RT @Paul__Walsh: 🛑 Do not add the ChatGPT Messages “plugin” to Apple Messages. And never use iMessage, SMS or RCS again. You’ll never know if the person you’re texting has been stupid and irresponsible enough to add this silent monitoring bot to your conversation. Governments no longer need to pressure Apple for a backdoor. OpenAI just built one for them. Every text you send them can end up on OpenAI’s servers, available to intelligence agencies and law enforcement through US legal process, and to other governments who can demand access through treaties and cross border agreements. Facebook Messenger is now safer than iMessage. @OpenAI just killed end to end encryption for Apple. Article in my drafts folder.

RT @Clashn0ut_: https://twitter.com/Clashn0ut_/status/2090546497158848895/video/1

RT @COTM_2: https://twitter.com/COTM_2/status/2090941868498415621/video/1

RT @deepseek_ai: DeepSeek-V4-Flash-Vision-Exp is now live on the DeepSeek API Platform! 🚀 🔹 This experimental multimodal model matches DeepSeek-V4-Flash on text capabilities—including agents, reasoning, and world knowledge. 🔹 On multimodal agent benchmarks, V4-Flash-Vision-Exp makes a major leap over V4-Flash, bringing multimodal agent performance close to Opus-4.8. Try it with model='deepseek-v4-flash-vision-exp'. DeepSeek Harness 0.1.1 was released today with out-of-the-box support for the new model. 1/n

RT @alessiapacca: be so fast that customers think you are a bot https://twitter.com/alessiapacca/status/2090582724541386956/photo/1

RT @DJIGlobal: Captured with DJI Mavic 4Pro and DJI Matrice 4TD, this dual perspective records a milestone moment — the launch and land recovery of the ZhuQue-3 Y2 rocket. * Filmed with authorization. Do not attempt. 🎥 ：LandSpace https://twitter.com/DJIGlobal/status/2090360385094168732/video/1

RT @ToonHive: ‘NIU LAI’ has reportedly dethroned ‘Obsession’ as the most profitable film of ALL TIME. The Chinese animated film has grossed over $4 MILLION on a budget of just $200. https://twitter.com/ToonHive/status/2090576946719555727/photo/1

RT @bilibili_create: We've just launched the intl app globally (currently on Android, iOS coming soon; and more countries coming soon) Globally distributed content, better localization & easier sign ups with no identity verification...and more localization optimization and features to come! Download the pink app with a globe on the top left for the international version! P.S. please join discord to report any issues or feedback

RT @bilibili_create: We've just launched the intl app globally (currently on Android, iOS coming soon; and more countries coming soon) Globally distributed content, better localization & easier sign ups with no identity verification...and more localization optimization and features to come! Download the pink app with a globe on the top left for the international version! P.S. please join discord to report any issues or feedback

RT @CtrlAltDwayne: Be Theo. Build an app. Customer uses app. Customer raises valid concerns about the product. Theo decides the correct response is to shit talk the customer on X, the everything app. Elite customer service. Well done, Theo. https://twitter.com/CtrlAltDwayne/status/2089911950482550811/photo/1

RT @AntoineRSX: i've said it. Niu Lai will be bigger than Odyssey. It's starting in china: https://twitter.com/AntoineRSX/status/2089902347598418189/photo/1

RT @AndrewCurran_: Detectors in ruins. This was always inevitable, one way or another. The moment models write at or above skilled human level they will be invisible to detectors, as the artifacts/patterns will no longer exist to detect. Good writing is just good writing. https://twitter.com/AndrewCurran_/status/2089846000454369589/photo/1

RT @alessiapacca: Whenever my schedule allows it, I love helping founders with YC interview prep. I remember how grateful we were to the YC alumni who gave us their time in the past. You just pay it forward.

RT @effectfully: how GitHub looks at its users after being forced to ingest 12 bajillion lines of hyperslop per second and then getting blamed for being down https://twitter.com/effectfully/status/2089392655419474388/photo/1

RT @alexinexxx: idc about the linux vs. mac os debate my only position is kill windows