OpenAI 发布 GPT-5.6 三模型矩阵
OpenAI 推出 GPT-5.6 系列三个层级模型：Sol（前沿）、Terra（均衡）、Luna（快速廉价），同时内部已全面用 Codex 智能体改造工作流。这是 OpenAI 首次用「模型家族」策略覆盖从高端推理到高频低价的完整需求光谱，直接对标 Anthropic 的 Claude 分层和 Google 的 Gemini 产品线。
OpenAI 这次命名很有意思：Sol/Terra/Luna 分别对应太阳、地球、月亮，暗示能力层级的同时也埋了「地月系」的扩展空间——未来可能还有 Mars 或其他行星代号。更关键的是内部 Codex 全面铺开这条信息，说明 OpenAI 不再只卖模型 API，而是在验证「模型即员工」的商业模式，这和 Anthropic 最近力推的 Computer Use 是同一战场。
LeCun 转发的 Mark Cuban 关于数据中心争议的推文被夹在一堆产品发布中间，暗示行业对算力扩张的政治反弹正在升温。做 AI infra 的团队需要关注美国各州的数据中心审批政策变化，这可能是比技术迭代更大的商业风险。
Introducing a limited preview of GPT-5.6 Sol, our next generation frontier model, as well as GPT-5.6 Terra, a balanced model for efficient, everyday work, and GPT-5.6 Luna, a fast and affordable model for high-volume work. https://t.co/OoM83SyISN
查看原文 →RT @mcuban: It’s time for everyone to realize that the fight against data centers has nothing to do with data centers. They have become a…
查看原文 →Work at OpenAI is being transformed by agents, in every department. Across our entire company, people are using Codex to do work that is more complex, longer-running, and increasingly cross-functional. Our internal usage offers an early look at how agentic tools may reshape work as they become more capable and broadly available.
查看原文 →RT @PabloTorre: Watching Kyra and her family fight brain cancer — glioblastoma — over the last year has been inspiring, and heartrending, b…
查看原文 →RT @Atul_Gawande: Independent analyses estimate that your actions to dismantle USAID and drastically reduce lifesaving foreign aid have alr…
查看原文 →RT @willccbb: something has definitely shifted in the past few weeks. seeing a huge uptick in large enterprises wanting to secure compute a…
查看原文 →How to build an agent that gets better over time: There are 3 areas an agent can learn from: 1. The model: Only works for code and math, where a computer can score right vs. wrong. Leave this to the big labs. 2. The harness: These are the steps, tools, and safety checks you build around the model. This is easy to control and will give you a huge payoff now. 3. The context: This is a plain-text representation of what the agent has learned. Probably the simplest place to start. But there's something else that most people miss: Your agent should learn from its users. You want to learn from every time a user fixes the agent's decision. Nothing can replace feedback from real usage.
查看原文 →RT @googlegemma: Gemma 4 just hit 200M downloads in only 2.5 months! For context, total downloads across the entire Gemma family of models…
查看原文 →RT @ihtesham2005: Elon Musk built one of the largest AI compute clusters on earth. Yann LeCun just explained why xAI now rents it out to ri…
查看原文 →RT @DegenerateTBone: Giant baby. Can't answer a few questions from the media about rainbow-colored hats, or take any accountability for his…
查看原文 →If you are a software engineer still using an IDE or a CLI to build software, you ain’t gonna make it. If you are writing prompts, or using skills, you ain’t gonna make it. Slack is the paradigm. That’s the only way forward. /s
查看原文 →The stickiness of Claude Tag will be like anything before. Claude Tag is designed to absorb every last bit of information from your company, and it will lock you out unless you keep paying them whatever they demand for eternity. You are paying Big AI to become hostage forever. Open source is the solution. Don't marry yourself to a company. Give yourself the flexibility to swap harnesses, models, and take your data wherever you want.
查看原文 →RT @ZcohenCNN: CNN reported in April that US intelligence assessed that roughly half of Iran’s missile launchers had survived US strikes.…
查看原文 →RT @RevantTeotia: 🎉 Our paper MJEPA has been accepted at #ECCV2026 !! Huge thanks to my awesome collaborators @AdrienBardes, @michaelrabba…
查看原文 →RT @williamswjt: @ZeffMax Because Dario sees himself as the high priest of the AI era, the only one qualified to stand between the most adv…
查看原文 →This agent relies on a marketplace of 2.8M+ specialized agents to complete a task. This is how I imagine the future of personal agents. Here is how this works: 1. Agentverse is a marketplace of agents 2. Anyone can build and publish an agent there 3. ASI:One is a personal agent with access to the marketplace 4. To solve a task, ASI:One finds the best agent for the job In other words: You have an agent that can solve pretty much *anything* because it knows how to outsource the work to a network of specialized agents. And you don't need to do any setup to get this working. The way ranking agents in the marketplace works is pretty cool: @Fetch_ai, the company behind it, created an AgentRank algorithm that works much like Google's PageRank. Every time an agent calls another, it forms a connection. Over time, these connections build into a graph of who relies on whom. An agent that is frequently called by other reputable agents scores higher and surfaces higher in results. The end result here is that your agent will consistently use the best agents in the marketplace to accomplish a goal. You should give this a try: https://t.co/r0F6dhl5kG. You don't need to pay or even register to try these agents. Thanks to the team for partnering with me on this post.
查看原文 →RT @randall_balestr: It's a bird, it's a plane, it's a JEPA! Congrats on that great work that brought SIGReg and JEPAs to the sky--in the r…
查看原文 →RT @PratyakshRao5: What should a world model for agile quadrotor control actually provide? 📄 Arxiv: https://t.co/fMzmm7GODZ 🌐 Project: ht…
查看原文 →RT @Dan_Jeffries1: There will be more and more pressure and more money and lobbying behind open soon. It is not a "two big companies have…
查看原文 →RT @KenRoth: More than 4.7 million people in the US have lost their Supplemental Nutrition Assistance Program benefits, also known as food…
查看原文 →