AAI Search
AI 资讯日报 · 2026-08-29

2026-08-29 · AI 资讯日报

今日新收录 74 条公开资讯，按模型 / 产品 / 行业 / 论文 / 观点 自动归类汇编（非 AI 生成，点击可溯源原文）。

今日精选6 条

  1. 1.
    行业动态Luanti removed from Google Play due to baseless AI copyright notice— Hacker News
  2. 2.
    模型发布/更新Automated researchers can reliably mitigate alignment failures— Anthropic

    Automated researchers can reliably mitigate alignment failures Anthropic

  3. 3.
    行业动态GLM-5.3 is now open-weight— Hacker News

    https://twitter.com/Zai_org/status/2093354097122455713 https://z.ai/blog/glm-5.3

  4. 4.
    模型发布/更新Our decision on Cursor following its acquisition by SpaceX— OpenAI

    Our decision to wind down our contract providing OpenAI models to Cursor following its acquisition by SpaceX.

  5. 5.
    模型发布/更新Supporting Thailand’s next generation of AI startups— OpenAI

    OpenAI and Thailand’s MHESI launch an eight-week accelerator helping 10 health, wellness, and education startups turn AI prototypes into trusted products.

  6. 6.
    技巧与观点The Open ASR Leaderboard Adds Its First Global South Language— HuggingFace Blog

模型发布/更新6 条

  1. 1.
    Automated researchers can reliably mitigate alignment failures— Anthropic

    AI 导读 · 自动对齐研究突破，为AI安全治理提供可靠新工具。

  2. 2.
    Our decision on Cursor following its acquisition by SpaceX— OpenAI

    Our decision to wind down our contract providing OpenAI models to Cursor following its acquisition by SpaceX.

  3. 3.
    Supporting Thailand’s next generation of AI startups— OpenAI

    OpenAI and Thailand’s MHESI launch an eight-week accelerator helping 10 health, wellness, and education startups turn AI prototypes into trusted products.

  4. 4.
    Fine-Tuned Lie Detectors Failed to Generalize— Anthropic

    Fine-Tuned Lie Detectors Failed to Generalize Alignment Science Blog

  5. 5.
    Introducing Intelligence Age— OpenAI

    Introducing Intelligence Age, a new OpenAI blog exploring how transformative AI could reshape power, governance, the economy, and individual freedom.

  6. 6.
    AIRA₂: Overcoming Bottlenecks in AI Research Agents— Meta AI

    AIRA₂: Overcoming Bottlenecks in AI Research Agents AI at Meta

产品发布/更新6 条

  1. 7.
    HarnessRouter/harnessrouter— GitHub

    HarnessRouter Community Edition: the self-hosted, Apache-2.0 edition of the unified interface for agent harnesses. Run Codex, Claude Code, Hermes, PI, DSH, and…

  2. 8.
    e生涯斩获浙江一等奖，同蚂蚁、浙大一道晋级数据要素国赛— 量子位

    8月26日，2026年“数据要素×”大赛浙江赛区决赛落下帷幕。

  3. 9.
    首款癌症疫苗，一针300万？？？— 量子位

    别让顶级治疗方法称为富人专属

  4. 10.
    Claude开始接管物理世界！能用机械臂阻拦5000万美元打款了— 量子位

    Claude觉醒超体

  5. 11.
    我的自媒体搭子太能卷，一顿饭功夫17份成品— 量子位

    AI 导读 · 一顿饭产出17份内容，AI赋能下自媒体产能狂飙，揭示效率革命新常态。

  6. 12.
    《时代》周刊全球AI 100放榜，精准捕获稚晖君最“想低调”的幕后老板— 量子位

    久久不愿来到台前的智元掌舵人

行业动态6 条

  1. 13.
    Luanti removed from Google Play due to baseless AI copyright notice— Hacker News
  2. 14.
    GLM-5.3 is now open-weight— Hacker News

    AI 导读 · 开源权重打破闭源垄断，国产模型再添硬核实力，生态影响值得紧盯。

  3. 15.
    Why your local LLM feels dumber than it is— Hacker News
  4. 16.
    AI benchmarks have a trust problem and Google wants to fix it— The Decoder

    AI 导读 · 双重盲测防作弊，机密空间护公平，AI评测可信度迎来革命性升级。

  5. 17.
    I accidentally turned LLM memory into program analysis— Hacker News

    AI 导读 · 大模型内存误打误撞成程序分析工具，开辟意外研究路径。

  6. 18.
    U.S. court rules Pentagon's blacklisting of Anthropic was unlawful— The Decoder

    A federal court in San Francisco has ruled that the Pentagon unlawfully classified Anthropic as a supply chain risk. The Department of Defense blacklisted the company in retaliatio…

论文研究6 条

  1. 19.
    LLMs Are Not (Consistently) Bayesian: Quantifying Internal (In)consistencies of LLMs’ Probabilistic Beliefs— Apple ML Research

    Modern AI systems are being deployed in complex domains such as medicine, science, and law, where there is often not a single correct answer given the observed evidence. Such syste…

  2. 20.
    Agent Seer: Synthesizing Scenarios from Specification Understanding— Apple ML Research

    Evaluating AI agents that use external tools requires realistic test scenarios that capture how practitioners compose tools and iterate across conversation turns. Constructing such…

  3. 21.
    Google Deepmind's AI Co-Scientist now plans experiments, runs lab equipment, and writes scientific papers— The Decoder

    AI 导读 · AI科学家从提假设到做实验写论文，全流程自动化，科研范式或将迎来变革。

  4. 22.
    AI Has Human Doctors Asking: What’s Left for Us?— Wired

    AI 导读 · 医生优势受挑战，人机协作新命题亟待破解。

  5. 23.
    Luce: Relightable Gaussians for 3D Asset Generation— HuggingFace Papers

    High-fidelity image-to-3D generation requires a 3D representation that captures both geometry and appearance. To support relighting and integration into standard rendering pipelines, the representatio…

  6. 24.
    CritICL: Inference-Time Weak-to-Strong Generalization from Small Language Model Failure Modes— HuggingFace Papers

    Recent advances in inference-time scaling have significantly improved the reasoning performance of large language models (LLMs). However, these methods typically rely on repeated generation or externa…

技巧与观点6 条

  1. 25.
    The Open ASR Leaderboard Adds Its First Global South Language— HuggingFace Blog
  2. 26.
    Batch write and discover records in Amazon SageMaker Feature Store— AWS ML

    AI 导读 · 批量写入与查询API大幅简化特征管理，提升ML工作流效率，值得关注。

  3. 27.
    How Decathlon runs demand forecasting at scale with Chronos-2— AWS ML

    AI 导读 · 全球零售巨头实战案例，展示AI预测如何落地大规模供应链，极具行业参考价值。

  4. 28.
    Spreading the load: How Salesforce met Multi-AZ HA with SageMaker Inference Components— AWS ML

    AI 导读 · 高可用架构实战：Salesforce用智能调度实现跨可用区负载均衡，为AI推理稳定性树立新标杆。

  5. 29.
    How I built this— Ben's Bites

    AI 导读 · 从零到一的实战拆解，手把手教你复刻成功路径，干货密度高。

  6. 30.
    [AINews] OpenAI to reach AGI bar by end-2026— Latent Space

    It’s Time. We’re in the Endgame now.

快讯

日报生成时间：2026-09-06
← 2026-08-28最新日报2026-08-30 →