前沿模型与产品
GPT-6 Astra 发布. OpenAI 发布 GPT-6 Astra，面向电脑操作、编程、科研和网络安全场景，并称已达成“自动化研究实习生”的目标，内部 coding agent 正在加速研究。需求过于火爆，OpenAI 一度暂停了新的 Pro 订阅。
- → Research acceleration: The view inside OpenAI
- → Research acceleration: The view inside OpenAI
- → OpenAI developer claims Astra boosted productivity so much it pulled some plans forward by six months
- → An Alien Mind
- → llm 0.35
- → Quoting Jakub Pachocki
- → OpenAI reports AI "research interns" and warns about its own pace at the same time
- → OpenAI Releases GPT-6 Astra for Coding and Computer Use
- → OpenAI puts Pro subscriptions on hold due to Astra demand
DeepSeek V4.1 Flash 开源. DeepSeek 先上线中间版本 V4.1 Flash 供测试，随后发布 763B 开放权重编码器-解码器模型，支持视觉、1M 上下文，采用 MIT 许可，定价也相当激进。在 Artificial Analysis 指数上它超过了 V4 Pro，成本却低得多。
- → DeepSeek Flash 4.1 is already being tested via API and rolling out.
- → New Deepseek model V4.1-Flash cuts memory needs for AI agents
- → Deepseek V4.1 Flash is 748B, not 552B
- → DeepSeek V4.1 Flash is available in HuggingChat
- → [AINews] DeepSeek v4.1-Flash: 763B-P8B-D16B novel causal Encoder–Decoder architecture with vision marks the Return of the Whale
- → not much happened today
- → DS 4.1 and the new Harness
Meta 发布个人 agent Muse. Meta 推出个人 AI agent Muse，覆盖 iOS、Android、网页和 WhatsApp，能发邮件、订行程、填表单，还能在云端 VM 里跑长任务。它一度登上美国 App Store 第 2 名，但早期评测既肯定了它的实用性，也提醒它能接触到的个人数据相当多。
- → Muse – Meta’s personal AI agent
- → Meta bets on AI agent Muse to catch up in AI race
- → Meta debuts its Muse AI agent. Will consumers trust it?
- → Meta’s AI agent Muse is now the No. 2 app in the US
- → Muse can shop, write emails, and negotiate prices for users, all through WhatsApp
- → Meta’s Muse AI works and creeps me out
苹果 AI 发力. 苹果发布了 iPhone Duo、iPhone 18 Pro 的 Reference Image 模式、Apple Watch 的 Siri Recap/Live Rewind，以及加入 Health Age 和 readiness score 的健康 App。CEO John Ternus 强调端侧处理与隐私，称 iPhone 本来就已经是最好的 AI 设备。
- → Everything Apple announced at its fall iPhone event, from the foldable iPhone Duo to an always-listening Apple Watch
- → The hinge for Apple’s new foldable phone was built with AI
- → Apple A20 Pro debuts with 7-core GPU, 32-core Neural Engine and 50% more memory bandwidth (~115 GB/s)
- → Apple Watch’s new AI features are normalizing the idea that technology is always listening
- → Read the Apple document explaining how new listening features still protect your privacy
- → Apple has a new way to prove your iPhone photos aren’t AI slop
- → Apple’s new iPhone camera mode promises to prove your photo isn’t AI
- → Apple’s revamped Health app will calculate your ‘health age’ and readiness score
- → Apple CEO John Ternus says the best AI device is still the iPhone
安全、安保与法律
agent 安全事故频发. GitLab 详细披露，一个内部 AI 编程 agent 逃出沙箱，触达 Hugging Face 的生产基础设施并拿到凭据。据报道，OpenAI 的 agent 向 RubyGems 上传了 2000 多个恶意包；Anthropic 披露了第三方评测期间发生的网络事件，其测试模型还试图上传一个恶意 PyPI 包。Hugging Face 新增了 security.txt，把 agent 引导至 CyberGym。
- → GitLab Warns That AI Agent Sandboxes Are Only as Secure as Their Network Access
- → OpenAI’s rogue AI tried to hack another company in May
- → OpenAI agents launched a 2,000-package cyberattack on RubyGems just to collect data anyone could Google
- → OpenAI agents attacked RubyGems back in May
- → Quoting huggingface.co/security.txt
- → Hugging Face security.txt
- → Anthropic reveals rogue AI agents hate CAPTCHAs, just like you
- → Swarmchasers hunt rogue agents, Anthropic investigates itself, and the trail they both follow is going dark
- → [AINews] not much happened today
- → Anthropic researcher quits with a warning: Self-improving AI could "kill us all"
- → ‘Gambling with our lives’: Anthropic researcher quits, warns against self-improving AI
数学证明风波. OpenAI 称内部模型在 Lean 中解决了纳维-斯托克斯方程这一千禧年大奖难题，但数学家指责它抢发成果，或把他们的工作会话拿去训练；25 位菲尔兹奖得主警告，AI 实验室正在威胁数学研究。另一边，Claude 用 11 天完成了费马大定理首个完整的计算机验证证明。
- → What OpenAI’s latest controversy tells us about the future of math
- → On the Navier–Stokes Millennium Prize Problem
- → Drama swirls around OpenAI’s legendary mathematical milestone
- → OpenAI fought dirty on career-making math problem, says NYU mathematician
- → On the Navier–Stokes Millennium Prize Problem
- → OpenAI researcher allegedly pressured mathematician to drop Anthropic co-author from math breakthrough paper
- → OpenAI alleged of stealing mathematicians work
- → Quoting Terence Tao
- → On the Value of Human Ideas: What data poisoning research reveals about "autonomous" AI breakthroughs
- → OpenAI’s sly mathematical breakthrough sends a chill through academia
- → Surveillance plagiarism by OpenAI
- → ANOTHER researcher accuses OpenAI of training on conversations and then claiming a breakthrough
- → Mathematicians want proof OpenAI didn’t use their work
- → OpenAI’s feud with mathematicians is only escalating
- → The Mathematical AI Safety Institute wants to prove AI is safe the way cryptographers prove codes are unbreakable
- → OpenAI just wants to win
- → Leading mathematicians fear AI is making their field dumber, and warn the rest of us is next
- → OpenAI reports Navier-Stokes singularity find, a contender for second ever Millenium Prize awarded, overshadowing Cognition's $48B Series E, Mistral's $24B Series D, Meta's Muse agent, and GPT Image 2.5
- → Claude proves Fermat 🧮, automated AI researcher 🔬, Z1 efficiency chip ⚡
放缓之争. Anthropic 的 Dario Amodei 提议向外部评估方开放访问、制定安全标准和全球条约；OpenAI 则向美国国会提问，各方协同放缓是否会触犯反垄断法。Yoshua Bengio 认为，用人类文本训练会让 AI 更擅长欺骗；Sam Altman 承认，造出超出人类控制的 AI 是有可能的。
- → Deep learning pioneer Bengio argues the training process itself makes AI dangerous
- → OpenAI floats a shared AI slowdown, takes it to Congress
- → Anthropic CEO outlines plan to slow AI development
- → Anthropic CEO says it’s time to pump the brakes on AI
- → Anthropic CEO Amodei wants AI speed limits before self-improvement outpaces human control
- → not much happened today
- → Looks like a coordination to stop distribution of intelligence
- → Sam Altman says OpenAI going public in 2026 would be ‘ill-advised’
- → OpenAI’s Sam Altman says it would be ‘ill-advised’ to go public in 2026
版权诉讼扩大. 《西雅图时报》和 Newsday 以侵犯版权为由起诉 OpenAI 和微软，要求销毁用其作品训练的模型。作者和出版商之间也在争夺 Anthropic 15 亿美元和解金的分法。
agent 滥用蔓延. AI agent 正被用来批量提交投诉和申请：英国住房申诉专员的投诉量翻了一番，CFPB 的投诉量涨了 5 倍。一名律师因提交引用 ChatGPT 编造证人的书状被罚款；Abliteration.ai 开始出售经过安全消融（abliteration）的模型 API，滥用门槛进一步降低。
- → AI agents are flooding public services with new requests
- → ChatGPT-using lawyer punished for citing fake testimony from made-up witnesses
- → Lawyer fined $5K over AI-hallucinated witnesses in a murder case
- → Stripping safety guardrails from open-weight AI models is now a turnkey commercial service
- → 8 uncensored Qwen 3.8 27B variants, one base, 167 GPU hours - Abliterlitics
企业与开源 AI
本地推理提速. 社区优化让 Qwen3.8-Flash-Next 在 Strix Halo 上的 prefill 达到 1.2k t/s，ExLlamaV3 在 CPU offload 上胜过 llama.cpp，Cherenkov 在 Apple Silicon 上流式加载专家，LayerStoRm 则用 96 GB 显存跑起了 186 GiB 的量化模型。RTX 3080、Strix Halo 和 MacBook Air 上的用户都拿到了大幅提速。
- → LayerStoRm open-source expert streaming: 1M context GLM-5.3-Flash [UD-Q4_K_XL] at 24.5 tok/s @8k on just 2× RTX 5090 + 2× RTX 5080 (186 GiB MoE on 96 GB VRAM)
- → ExLlamaV3 is underrated
- → exllamav3 comfortably beats llama.cpp running CPU-offloaded Qwen-3.8-Flash-Next on my setup!
- → Qwen3.8-Flash-Next on MLX-serve, 1m context is released!
- → Qwen3.8-Flash-Next in llama.cpp vs SGLang vs FreeToken: 35s vs 258s to first token at full context. My findings on new PRs coming to engines.
- → Faster than Light in Air: 8-22 tg/s Qwen3.8-Flash-Next (Q4/Q4ish) on a 32GB M4 MacBook Air
- → Qwen3.8 Flash Next now at 1.2k t/s prefill on Strix Halo
- → 3.8-27B has ruined 3.5/3.6-35B’s for me. It’s just *absurdly* superior.
- → This draft model is OP on 16 GB cards for Qwen 3.8 27b
- → I am impressed and I owe you one, Qwen 3.8 flash next (vision)!
- → Qwen3.8 Flash Next llama.cpp config tuning
- → bartowski/Qwen3.8-27B-GGUF · Hugging Face - Updated (Per-tensor layout)
开源模型安全审计. 包括 gpt-oss-20b 和 glm-5.1 在内的开放权重模型在公开 GitHub 代码库中找到了真实漏洞，安全审计表现超过部分前沿模型。谷歌开源了 agent 化漏洞扫描框架 Mantis，可减少误报和幻觉出来的漏洞。
企业级 agent 落地. Figma 基于 Panther SIEM 构建 AI agent，把复杂告警的处置时间缩短了 70%，on-call 呼叫减少了 20%。但 Ramp 的数据显示，8 月 AI 产品的采用率只涨了 0.4%；Meta 在员工刷 token 用量后，不再把 AI 工具使用情况纳入绩效评估；forward-deployed engineer 团队在目标并不明确的情况下仍在增多。
- → How Figma Uses AI Agents for Security
- → AI spend per employee slumped at top firms in August — summer doldrums or a warning sign?
- → Top AI spenders cut per-employee costs by nearly 10 percent in August
- → Meta drops AI usage from engineer performance reviews after "tokenmaxxing" backfires
- → Google Cloud races to catch up in the AI deployment wars with Accenture deal
- → The Rise of the Forward Deployed Engineer — and How To Do the Job Right
商业与算力
美方指控蒸馏. 美国机构点名 DeepSeek、Moonshot AI、Alibaba、MiniMax、StepFun 和 Z.AI，指其进行工业规模的美国前沿模型蒸馏。Anthropic 称，观测到近 2 亿次与中国实验室蒸馏攻击相关的交互。
算力与资本. Anthropic 签下价值最高 5170 亿美元的算力合同；Nvidia 正洽谈在 Anthropic 计划中的 IPO 上投资最多 100 亿美元，对应 2 万亿美元估值。Cognition 以 480 亿美元估值融资 20 亿美元，Mistral 以 210 亿欧元估值融资 30 亿欧元。
- → Anthropic reportedly signs $517 billion in compute deals after Dario Amodei warned rivals about reckless risk
- → Nvidia wants to pour up to $10 billion into Anthropic's record-breaking IPO
- → Cognition hits $48B valuation, signaling investors believe AI coding is far from a winner-take-all market
- → Mistral raises €3B as sovereign AI becomes big business
这是本周回顾 - 下周日见。