今日必看

Set an expiration date for individual user budgets

You can now set an optional expiration date on an individual user budget, and GitHub removes the budget on that date. This capability is generally available. The user then falls ba

Show HN: AI coach that adapts workouts and meals to your injuries and

Show HN: Dehurdle – Real-Time AI Voice Simulations for High-Stakes Conversations

Show HN: Every AI agrees with you. This writes your startup's obituary instead

Show HN: I put four persistent AI minds in a room and let them change each other

Show HN: Redactle LLM Leaderboard

I got nerd sniped and ran a bunch of evaluations to see how well models could play Redactle. It turns out Gemini flash is in a league of it's own in both performance and cost. It's

Show HN: StealthOS, an iOS privacy sandbox with local AI and Tor

Show HN: The Drudge Report, but for AI

Skild AI 发布旗舰机器人基础模型 S1

The Robot Report 报道 Skild 推出 S1：面向真机技能的机器人基础模型，定位 physical AI 的通才策略。

SkillGLoW：长程技能巩固

按程序家族巩固技能，自改进长程 agent。

Small and medium teams

SolarWM: Open Data and Scalable Training for Long-Horizon Video World Models

We introduce SolarWM, a fully open foundation for building interactive video world models from data preparation through long-horizon inference. Training across heterogeneous data s

Sonos Ace Ultra, Beam Ultra, Sonos Fabric, and a New App: Everything Sonos Just Announced

Sonos is cramming AI into its software because it’s “very hot these days.” The new features, which include agentic automation, are opt-in.

Spectral Initialization and Scheduled Graph Smoothness for Uncertain Knowledge Graph Completion

Uncertain knowledge graphs (UKGs) extend knowledge graphs by assigning each triple a continuous confidence score. Since most possible triples lack observed confidences, recent meth

StudentSim：训练学生模拟器

用 LLM 训学生模拟器。

Supporting Thailand’s next generation of AI startups

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

TaRA: Training-Aware Low-Rank Adaptation Initialization

Low-Rank Adaptation (LoRA) has become a de facto standard for parameter-efficient fine-tuning (PEFT), yet its performance is highly sensitive to initialization due to the informati

Team 计划 Copilot 模型访问更新

8 月底 changelog，仍在窗口内。

Terminal-Bench-Science 0.1：评估科研工作流中的 AI 智能体

斯坦福大学研究人员领衔发布 Terminal-Bench-Science 0.1，用来自生命、物理、地球、数学和工程科学的 70 个专家精选任务评估 AI 智能体的科研能力。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmtcxdp3f07l3roq5uk6om1aw

Terminal-Bench-Science 0.1：评估科研工作流中的 AI 智能体

斯坦福大学研究人员领衔发布 Terminal-Bench-Science 0.1，用来自生命、物理、地球、数学和工程科学的 70 个专家精选任务评估 AI 智能体的科研能力。

The Builders Stage brings practical strategies for scaling startups to TechCrunch Disrupt 2026

The Builders Stage is returning to TechCrunch Disrupt, bringing together founders, startup operators, and investors for practical conversations on what it takes to build and scale.

The Cybersecurity Apocalypse Is Coming in ‘Months,’ AI Giants Warn

Plus: Hackers target over 100 US water systems, ICE puts in an order for robot dogs, and you’ll never guess what “MrChildPorn” was arrested for.

The Decoder：Omni 1.1 Flash 让 AI 视频更便宜更灵活

媒体称 Google 新 Flash 视频模型在价格和镜头控制上同时推进，适合更长镜头与后期放大。

The Decoder：Runway Solaris 实时生成软件界面

媒体解读 Runway Interface World Model 首作 Solaris：逐帧实时生成可交互界面，而不只是预渲染视频。