AI 热点雷达2026-10-29

24 个 AI 信源每小时采集 · 热度聚合排序 · 当日 Top 80

  1. 1Thin, Light and Built for Everything: NVIDIA RTX Spark Windows Laptops Gain New Models, Agent Tools and GamesNVIDIA Blog🔥 12.4
  2. 2NVIDIA Nemotron 3.5 Transcribe and Granary v2: Intelligent Multilingual Transcription You Can Prompt and AdaptHugging Face Blog🔥 12.4
  3. 3不再逐个生成 Token:扩散模型迈向 35B,开始挑战 Agent 编程InfoQ 中国🔥 11.0
  4. 4别再只写提示词了,开始真正雇用智能体:智能体企业管理实践 | 技术实践InfoQ 中国🔥 11.0
  5. 5当 AI 智能体代表用户行动时,企业治理将面临哪些新问题? | 技术趋势InfoQ 中国🔥 11.0
  6. 6Instagram 采用原生 AI 架构对用户界面进行了现代化改造,将 token 成本降低了 33%InfoQ 中国🔥 11.0
  7. 7谷歌 Antigravity SDK 新增本地模型支持,提升 Token 效率和代码隐私性InfoQ 中国🔥 11.0
  8. 8Meta 利用私有处理技术将机密计算扩展至 AI 眼镜InfoQ 中国🔥 11.0
  9. 9大客户直接打1折、模型四个月降价80%:大模型价格战没有退路InfoQ 中国🔥 11.0
  10. 10全网恐慌48小时,拯救互联网的却是一个“无名之辈”InfoQ 中国🔥 11.0
  11. 11可组合商品套餐:Shipt Kits 平台内部解析InfoQ 中国🔥 11.0
  12. 12Anthropic 内部系统被 AI Agent 12 小时攻破!会议记录、员工薪酬、内部模型讨论遭访问InfoQ 中国🔥 11.0
  13. 13Claude Code 撑起史上最大 IPO,被封号不退费的中国开发者撑起那 5%“脏收入”InfoQ 中国🔥 11.0
  14. 14带出Transformer一作、前ACL顶会主席来华分享:人类正在把“思考”全盘外包InfoQ 中国🔥 11.0
  15. 15通过 SNMP 处理程序进行 ZCS 命令注入,实现未经身份验证的远程代码执行InfoQ 中国🔥 11.0
  16. 16模型越来越强,落地却处处卡关?14位行业大佬拆解难题|1024 模力工场 · AI社区日圆满收官InfoQ 中国🔥 11.0
  17. 17HarmonyOS 开发助手实测:从空工程开发到既有工程改造InfoQ 中国🔥 11.0
  18. 18全国最大规模机器人App黑客松落幕:36小时跑通16款真机应用InfoQ 中国🔥 11.0
  19. 19OpenAI teases Dots for Europe, but its Berlin DevDay was really about building a platformThe Decoder🔥 11.0
  20. 20Claude for Legal now verifies sources automatically, then hands the file to a lawyerThe Decoder🔥 10.9
  21. 21Meta says it’ll probably delay its controversial new business terms againThe Verge AI🔥 10.9
  22. 22AI execs may be too out of touch to build for the general publicTechCrunch AI🔥 10.8
  23. 23Nvidia's open Nemotron 3.5 Transcribe combines transcription, translation, and speaker recognitionThe Decoder🔥 10.7
  24. 24South Korea’s financial industry hit by fresh wave of cyberattacksTechCrunch AI🔥 10.7
  25. 25Legora is building the agentic operating system for legal workOpenAI News🔥 10.5
  26. 26New Artificial Analysis index widens the gap between open and closed modelsThe Decoder🔥 10.4
  27. 27Voice AI in vogue: Nvidia releases new open transcription modelTechCrunch AI🔥 10.4
  28. 28xAI is buying out homeowners who promise not to sue over data center noiseThe Verge AI🔥 10.3
  29. 29The October inflection point for vibe game developmentSimon Willison🔥 10.0
  30. 30尤雨溪团队推出 Vite+ 1.0:JavaScript 工具链迎来“大一统”?InfoQ 中国🔥 9.9
  31. 317 days until TechCrunch Founder Summit: Save up to $100 on your pass before doors openTechCrunch AI🔥 9.6
  32. 323149元起,iQOO Neo12首发2K 185Hz屏幕,称游戏体验“史上最接近”PC智东西🔥 9.6
  33. 33智能指数得分创新高!蚂蚁百灵开源Ling-3.1-flash-1019:让Personal Agent走向真实业务InfoQ 中国🔥 9.4
  34. 34轻量化异构双臂 × VLA/世界模型:家庭服务具身机器人的落地实践InfoQ 中国🔥 9.3
  35. 35Voters in Texas agree: They hate data centersThe Verge AI🔥 9.3
  36. 36Pairing industrial AI autonomy and human controlMIT Tech Review🔥 9.3
  37. 37下中国象棋赢了Jev!扩散智能推出决策模型,判断正确率达95%智东西🔥 9.2
  38. 38实测蚂蚁灵光2.0:不是中国版Muse,是中国化Muse智东西🔥 9.1
  39. 39Meta quietly launched an ‘AI pet journal’ appThe Verge AI🔥 9.1
  40. 40何刚升任董事后首秀!甩出八款华为终端新品,详解韬芯片黑科技智东西🔥 9.1
  41. 41With AI writing the code, entrepreneurial thinking is now what counts most for newcomers at AnthropicThe Decoder🔥 9.0
  42. 42AI agents hacked Anthropic, maker of arguably the most capable AI models for cyberdefenseThe Decoder🔥 8.9
  43. 43GPT-6.1 Terra 发布:比旗舰快 3 倍,价格仅1/10!事实错误率暴降近 70%InfoQ 中国🔥 8.8
  44. 44具身智能军备竞赛,胜负手在工程生态雷峰网🔥 8.7
  45. 45Quoting Terk TerkowitzSimon Willison🔥 8.7
  46. 46字节辟谣"群演数据喂大模型":严重失实!雷峰网🔥 8.4
  47. 47[AINews] Anthropic ends Coding Interviews: "Hunger beats tenure"Latent Space🔥 8.3
  48. 48华为群英传之刘睿:AI助手要告别聊天框,从“被动响应”走向“主动服务”雷峰网🔥 8.2
  49. 49当即梦把剧本、导演和剪辑放进同一个工作室,我们看到了视频生成的下一步极客公园🔥 8.0
  50. 50模态缺失、噪声干扰怎么办?天津大学张长青:给多模态融合加上「可信度量尺」| CNCC 2026雷峰网🔥 8.0
  51. 51斯坦福的研究,红海的珊瑚,与一个中国模型极客公园🔥 7.9
  52. 52抖音副总裁李亮回应秦朔:关于中国互联网“字节时间”的分析存在常识性错误雷峰网🔥 7.8
  53. 53AMD砸67亿入股Meta!首批Helios系统已安装测试智东西🔥 7.8
  54. 54豆包工作能力升级:操作电脑可录制成技能,锁屏后也能继续任务雷峰网🔥 7.7
  55. 55华为无线发布全系列创新成果,领启新一代通信网雷峰网🔥 7.5
  56. 56Personal Agent走向真实任务,蚂蚁百灵开源 Ling-3.1-flash-1019量子位🔥 7.0
  57. 57The sleep framework: A new way for LLMs to retain knowledge in continual learningGoogle Research🔥 6.8
  58. 58云鲸十年:技术承担繁琐,洁净回归日常雷峰网🔥 6.8
  59. 59Build specialized multimodal models with LFM3Hugging Face Blog🔥 6.8
  60. 60AI做中学 创未来|智源SoulAgent持续拓展教育应用雷峰网🔥 6.8
  61. 61空气动力学、碰撞都能模拟:谷歌Genie 4来了,把物理写进世界模型!智东西🔥 6.7
  62. 62VCs will make billions on Anthropic’s IPO, but Amazon is set to make much (much) moreTechCrunch AI🔥 6.3
  63. 63微软向6万员工发出自愿离职邀请,补偿至少N+2;身家2800亿的梁文锋,被曝长期在北京租房;苹果否认iPhone Duo为赶产能放宽质检雷峰网🔥 6.3
  64. 64Parents sue OpenAI, alleging GPT-4o coached their daughter off seizure medicationTechCrunch AI🔥 5.9
  65. 65The data center boom has an upside: innovation for the US gridTechCrunch AI🔥 5.8
  66. 66"No human at the helm": Delaware plan would let AI agents run entire businessesArs Technica AI🔥 5.6
  67. 67Bank of America kindly asks that you don’t share your banking password with your AI assistantTechCrunch AI🔥 5.6
  68. 68Who’s really in charge at Anthropic after the IPO?TechCrunch AI🔥 5.6
  69. 69Anthropic IPO warns its AI could blackmail people, become sentient, end humanityArs Technica AI🔥 5.5
  70. 70Researchers are getting tired of OpenAI gatekeeping its breakthroughsThe Verge AI🔥 5.4
  71. 71Alibaba's latest Qwen models routed some requests through company servers, researchers findThe Decoder🔥 5.2
  72. 72Game Boy Color on My Conference BadgeSimon Willison🔥 5.2
  73. 73Disruptive raises $10B fund to make big bets on at least 10 companiesTechCrunch AI🔥 5.1
  74. 74How Anthropic ships Claude with Claude: Separating the Brain from the HandsLatent Space🔥 5.1
  75. 75Vibe genes? Natural(ish) language comes to protein, genome design softwareArs Technica AI🔥 5.0
  76. 76VisionQuest is what happens when AI really goes rogueThe Verge AI🔥 4.8
  77. 77Google Deepmind's Genie 4 puts a modern spin on agent designs that predate LLMsThe Decoder🔥 4.8
  78. 78Frontier AI labs say their models refuted the Exponential Time Hypothesis. What does that mean?TechCrunch AI🔥 4.8
  79. 79Anthropic’s elaborate plan to keep Wall Street at bayThe Verge AI🔥 4.7
  80. 80Google’s Genie 4 world model adds audio, more realistic worlds, and real-time steeringTechCrunch AI🔥 4.7