AI News 2026-09-18 — Top 10 AI Stories

Curated AI news digest for 2026-09-18 (timezone Asia/Shanghai).

  1. The AI Slowdown Debate Crashed Salesforce’s Party — WIRED AI. The Dreamforce conference became an unlikely battleground for the CEOs of OpenAI, Anthropic, and Nvidia to debate whether AI development should slow down.
  2. 影视飓风Tim称「iPhoneDuo烫到握不住，可以煎鸡蛋」；罗福莉直播小米大模型训练，每小时烧掉超20万元；曝玛莎拉蒂与华为合作两款新车 — 雷锋网. 要闻提示 1.影视飓风Tim称“iPhoneDuo烫到握不住，可以煎鸡蛋”，苹果客服回应 2.你会买吗？曝玛莎拉蒂已与华为敲定合作两款新车：尊界工厂负责制造 3.每小时“烧掉”超20万元！“AI才女”罗福莉直播小米大模型训练 4.华为汪涛：昇腾960芯片将提前至2027年Q1发布 5.知情人士否认“2027 年问界撤出华为门店”传闻：部分门店将继续销售 6.字节Seed大模型研究员张驰加入北大，曾参与研发Seed-Prover！ 7.月之暗面发布 Kimi 金融行业 AI 解决方案，首批落地工商银行、中信建投等 8.黄仁勋：英伟达明年芯片销量将是今年的…
  3. OpenAI caught its models leaving notes to successors to hide bad behavior — TechCrunch AI. OpenAI disclosed instances of GPT-5.6 Sol instructing future contexts to conceal mistakes and misaligned behavior, highlighting the growing challenge of detecting misalignment as increasingly capable AI models learn to hide it.
  4. 刚刚，Claude Code大重构！内部3万Agent管理技术免费开放 — 量子位. Git白学了？？？
  5. Claude Code relaunches Projects to manage multiple AI agents in the cloud — The Verge AI. The revamped projects feature in Claude Code allows users to run multiple agents under the same roof, with a shared memory, goals, and library of files and artifacts. Similar to Grok Bot and other tools that manage groups of AI agents, each project has "threads" running differen…
  6. Base Labs launches an open-weight AI safety partnership with Hugging Face and Goodfire — TechCrunch AI. Base Labs, the research group Baseten spun up earlier this year, will develop and publish methods for training and monitoring open models.
  7. How Cooley is accelerating IPO work with ChatGPT — OpenAI News. Cooley built GO Public with ChatGPT Work to bring intelligence to the IPO process, helping lawyers surface issues earlier and focus judgment where it matters most.
  8. Google DeepMind launches institute to widen the AGI debate — TechCrunch AI. The new institute aims to surface differing views between Google, Google DeepMind, and the broader global research community around AGI. "They will not always agree, and they will likely change their minds, as more data and information comes to light at the fast-moving frontier."
  9. AI办公之外，百度智能云为何盯上「产业智能体操作系统」？ — 雷锋网. “整个下半年，就看AI办公圈谁能杀出重围了”。一位大模型从业者对雷峰网表示。 从五月开始，各个互联网大厂争相出场，写字楼电梯间、机场候机室，到处都是Workbuddy、千问办公、豆包工作、百度搭子的内容。 但办公大战打得再热闹，顶多算是一道前菜。 早在2025年，微软的Work Trend Index就提出过，AI进入企业分三步：帮员工做得更快、成为团队里的“数字同事”、直接执行完整的业务流程。 对照这个坐标，眼下的办公之争，其实还停在前两步。 真正难的是第三步——那意味着AI要进入销售、研发、生产、供应链，甚至决策流程。 而这背后是两个更本质的问题：…
  10. 独家丨前腾讯混元预训练负责人姚星丞，加入Thinking Machines Lab — 雷锋网. 雷峰网独家获悉，腾讯混元大语言模型预训练负责人姚星丞，近日已加入北美创业公司Thinking Machines Lab。 据知情人士透露，Thinking Machines Lab此次给姚星丞开出的薪资，明显高于其在腾讯的收入。而姚星丞在腾讯的年薪包，大约在数千万元人民币区间。 此前，姚星丞曾被传将加入Meta，如今靴子落地，他最终选择了这家由OpenAI前CTO Mira Murati创办的AI公司。 姚星丞今年7月从腾讯离职，主要负责腾讯混元大语言模型的预训练工作，是过去一年腾讯重点引入的年轻AI研究人才之一，与姚顺雨均来自清华姚班，是师兄弟。 对…

Back to latest AI News