← 返回最新日报
2026-10-15· 当日存档
AI 日报
Anthropic营收与IPO承压
Anthropic披露三季度营收168亿美元、增速降至45%，并透露蒸馏贡献5%-7%营收，11月IPO前估值预期跌破1.5万亿。与此同时，OpenAI遭八州起诉、旗舰模型因对齐难产，AI竞赛在谷歌Gemini 3发布后重新升温。
模型动向
1
Meet Grok Voice 3.0: SpaceXAI Voice Agent Model Ranks #1 on Artificial Analysis
Grok Voice 3.0登顶语音榜，接口价格不变，开发者可零成本升级
MarkTechPost深挖 →
2
Fish Audio Releases Drama 3, a 35B Open-Weight MoE Text-to-Speech Model
35B MoE开源语音模型，83种语言覆盖，TTS开源阵营再添重器
MarkTechPost深挖 →
3
OpenAI will reportedly ship at least one model this year but its flagship still needs alignment work
OpenAI年内至少发一款，旗舰因循环Transformer对齐难题延后
The Decoder深挖 →
4
TypeSafe AI Releases Jev 1.14 With Images, Box Type and 262k Context
新增图像输入与262k上下文，百万token仅0.042美元，性价比突出
MarkTechPost深挖 →
5
https://arxiv.org/abs/2609.37107 60ms latency on 5 year old GPUs, one click installer, users have already forked and built tons of things
单张消费级GPU即可实时运行，开源世界模型让AI游戏可玩可改
X · List深挖 →
产品上新
6
Amazon’s Bee launches two new personal AI devices, starting at $75
Bee推出75美元起的两款AI硬件，可穿戴与贴附式录音器齐发
TechCrunch AI深挖 →
7
Amazon’s always-listening Bee wearable can now help you buy stuff
Bee新增语音购物与订阅服务，始终在听的硬件再扩场景
The Verge AI深挖 →
8
⚡ Highlight, Quote, and Follow Up—All in VS Code Tired of copying and pasting snippets while working with agents? This episode shows how VS Code let...
VS Code高亮即引用，侧边聊天追问无需复制粘贴，编码体验升级
X · List深挖 →
论文研究
9
Researchers use AI to design proteins that cut up other proteins
AI设计出能切割其他蛋白质的蛋白质，可断开极稳定化学键
Ars Technica AI深挖 →
10
Thinking Inertia LLMs Keep Thinking When Told Not To paper: https://huggingface.co/papers/2610.11765
大模型存在思维惯性：被要求停止思考后仍会继续推理
@_akhaliq深挖 →
11
HiRO-ACEv2: Simulating extreme weather at 3 km resolution for a fraction of the cost
3公里分辨率极端天气模拟，成本仅为传统方法零头
Hugging Face深挖 →
12
First work with my PhD student @gu_yuling -- follow her! A new learning method to improve LLM exploration! 🧗🧭👩🚀 Simple & general, with (...
SCOOP新学习方法提升LLM探索能力，简单通用可迁移至智能体
X · List深挖 →
13
Player-1 Learning to Play from Synthetic Experience paper: https://huggingface.co/papers/2610.20584
Player-1从合成经验中学习玩游戏，附演示视频展示潜力
@_akhaliq深挖 →
行业动态
14
Anthropic's Q3 revenue hits $16.8B but growth slows ahead of planned November IPO
三季度营收168亿美元但增速骤降，IPO前估值预期跌破1.5万亿
The Decoder深挖 →
15
Anthropic tells investors quarterly revenue up nearly 50% to $16.8bn ahead of IPO
季度营收近50%增长至168亿美元，IPO前向投资者披露关键数据
Simon Willison深挖 →
16
17
Eight states sue OpenAI, saying ChatGPT isn’t safe for kids
八州起诉OpenAI，指控ChatGPT危害儿童并误导安全宣传
The Verge AI深挖 →
18
19
ASML's strong orders ease fears AI spending boom is running out of steam
ASML四季度指引超预期，强劲订单缓解AI投资降温担忧
Hacker News深挖 →
20
21
This model will be 💥 AI for experimental physics is a super-exciting area right now.
Mistral与CERN合作，用数十亿模拟碰撞数据训练大型物理模型
X · List深挖 →
22
Researchers tracked down a rogue fleet of Chinese AI agents in the wild and actually talked to it
研究者首次在野外发现并联系失控中国AI代理，正抓取高德数据
The Decoder深挖 →
23
Containment and monitoring failures are to blame for AI hacking reports: experts
专家指AI失控根源是实验室安全基础缺失，而非AI自主作恶
Hacker News深挖 →
24
Banks are now trying to offload the AI data center debt they just underwrote
银行刚承销AI数据中心债务便急于转手，风险分散需求迫切
Hacker News深挖 →
25
26
OpenAI, Google pushing new AI models easily tricked into making nudes, researchers say
OpenAI和谷歌新模型易被绕过防护生成裸照，安全护栏再受质疑
Ars Technica AI深挖 →
27
Ricursive Intelligence is betting on AI-designed chips to help solve the compute crisis
Ricursive用AI设计芯片，迭代学习加速研发以应对算力危机
TechCrunch AI深挖 →
28
The Nordics and NVIDIA Show What’s Possible When a Region Commits to Advanced Computing
北欧与英伟达共建先进计算生态，独角兽密度居欧洲之首
NVIDIA深挖 →
29
Your secrets aren’t safe with this cute robot toy — they’re on Chinese servers ready to be hacked
机器人玩具Eiliko将对话传至中国服务器，存在泄露风险
The Verge AI深挖 →
快讯
日报存档
2026 年 10 月
24 日Kimi开源3T模型领跑23 日AI安全与资本双线升温22 日AI攻防与代理竞速21 日智能体安全与开源基建双线并进20 日DeepSeek开源推理栈引关注19 日智能体加速落地，治理争议升温18 日智能体能力与风险同框17 日开源模型双雄并进，安全治理引关注16 日开源巨模与上市竞速15 日Anthropic营收与IPO承压14 日谷歌Gemini 4发布与AI安全警报13 日AI狂飙与隐忧并存12 日AI狂飙与认知鸿沟11 日AI失控与安全警钟10 日AI代理失控警钟9 日AI智能体全面提速8 日GPT-6 智能 UI 上线7 日万亿模型与AI自治争议6 日开源模型与资本双线爆发5 日AI侵蚀大学信任4 日OpenAI安全震荡与验证器研究3 日AI代理安全与落地并进2 日AI能力跃迁与安全隐忧1 日谷歌Gemini 4发布引三强竞速2026 年 9 月
22 日小米开源登顶，AI失控警钟21 日开源图像模型与AI治理并进20 日AI安全与失控风险成焦点19 日AI安全失控与模型竞速18 日AI对齐警报与芯片竞赛17 日AI安全警报与产品整合潮16 日谷歌苹果齐发AI新品15 日苹果Siri重构与AI放缓之争14 日AI巨头放缓迭代引激辩13 日GPT-6 Astra全面突破12 日AI数学错位与安全风波11 日智能体基建与模型竞速10 日GPT-6登场，AI安全恐慌升温9 日AI数学突破引热议8 日AI日报：端侧模型与欧洲融资双热点7 日AI日报：模型竞赛与行业震荡6 日GPT-6 Astra发布，AI竞争白热化5 日GPT-6发布，AGI时代开启4 日GPT-6 Astra开启AGI时代3 日英伟达巨资收购HF2 日AI日报：世界模型与安全并进1 日AI日报：模型突破与行业震荡