TypeScript · ★ 122,827 · 🍴 6,590 · 📈 2,650 stars today
Collection of publicly available IPTV channels from all over the world
中文介绍 全球公开 IPTV 频道资源聚合项目，收录来自世界各地数千个可免费收看的直播频道，以 M3U 播放列表格式组织，涵盖新闻、体育、娱乐等分类。适合希望在各类播放器或智能电视上收看国际频道的用户使用，无需付费订阅。
Elixir · ★ 8,223 · 🍴 946 · 📈 35 stars today
A self-hosted data logger for your Tesla 🚘 [main maintainer=@JakobLichterfeld]
中文介绍 面向 Tesla 车主的自托管数据记录工具，自动采集并持久化车辆行驶数据，包括里程、能耗、充电记录、地理位置等，通过 Grafana 仪表板可视化展示。基于 Elixir 开发，配合 PostgreSQL 存储，适合注重数据隐私、希望深入分析自己 Tesla 使用习惯的车主部署在家庭服务器上。
Python · ★ 30,085 · 🍴 2,446 · 📈 1,045 stars today
Give your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
中文介绍 为 AI Agent 提供「互联网之眼」的 CLI 工具，支持读取和搜索 Twitter、Reddit、YouTube、GitHub、Bilibili、小红书等主流平台内容，无需申请各平台 API Key，零费用即可使用。适合开发者将其集成到 AI Agent 工作流中，实现跨平台信息获取与自动化信息检索场景。
TypeScript · ★ 10,622 · 🍴 3,434 · 📈 227 stars today
Meshery, the cloud native manager
中文介绍 云原生基础设施管理平台，定位为统一的云原生管理器，支持对 Kubernetes、服务网格（如 Istio、Linkerd）及相关云原生组件的可视化管理、配置与性能测试。CNCF 沙箱项目，适合 DevOps 工程师和平台团队在多集群、多云环境下统一管理云原生工作负载。
Ruby · ★ 31,665 · 🍴 7,628 · 📈 431 stars today
Open-source live-chat, email support, omni-channel desk. An alternative to Intercom, Zendesk, Salesforce Service Cloud etc. 🔥💬
中文介绍 开源全渠道客户支持平台，整合实时在线客服、邮件、社交媒体、WhatsApp 等多种沟通渠道于统一工作台，可作为 Intercom、Zendesk 的自托管替代方案。基于 Ruby on Rails 与 Vue.js 构建，适合中小企业搭建自有客服系统，掌控客户数据并降低 SaaS 订阅成本。
Java · ★ 126,874 · 🍴 15,135 · 📈 95 stars today
《Hello 算法》：动画图解、一键运行的数据结构与算法教程。支持简中、繁中、English、日本語，提供 Python, Java, C++, C, C#, JS, Go, Swift, Rust, Ruby, Kotlin, TS, Dart 等代码实现
中文介绍 《Hello 算法》是一本以动画图解和可直接运行代码为核心特色的数据结构与算法入门教程，支持简繁中文、English、日文四种语言，提供 Python、Java、C++、Go、Rust 等 14 种编程语言实现。适合编程初学者系统学习算法，也适合备战技术面试的开发者查阅参考。
TypeScript · ★ 447,865 · 🍴 44,989 · 📈 738 stars today
freeCodeCamp.org's open-source codebase and curriculum. Learn math, programming, and computer science for free.
中文介绍 freeCodeCamp.org 的开源课程与平台代码库，提供免费的编程、数学和计算机科学学习路径，涵盖 Web 开发、数据科学、机器学习等方向，并附带认证体系。面向全球希望零成本自学编程的学习者，已有数百万人通过该平台完成技能转型。
HTML · ★ 18,134 · 🍴 1,171 · 📈 57 stars today
Open-source infrastructure for Computer-Use Agents. Sandboxes, SDKs, and benchmarks to train and evaluate AI agents that can control full desktops (macOS, Linux, Windows).
中文介绍 面向 Computer-Use Agent（计算机操控型 AI）的开源基础设施，提供沙箱环境、SDK 及基准测试套件，支持训练和评估能够操控完整桌面系统（macOS、Linux、Windows）的 AI Agent。适合 AI 研究者和工程师开发、测试具备 GUI 交互能力的自动化 Agent。
★ 352,284 · 🍴 83,563 · 📈 352 stars today
A complete computer science study plan to become a software engineer.
中文介绍 一套系统完整的计算机科学自学计划，专为目标进入大型科技公司（如 Google、Amazon）的软件工程师面试准备而设计，涵盖数据结构、算法、操作系统、网络等核心知识点，并附有学习资源推荐和打卡追踪表。适合非科班转行或需要补充 CS 基础的开发者长期使用。
Python · ★ 33,059 · 🍴 5,408 · 📈 538 stars today
Learn it. Build it. Ship it for others.
中文介绍 从零开始学习 AI 工程的实践课程仓库，强调「学了就做、做了就发布」的闭环理念，帮助开发者系统掌握 AI 应用工程化落地所需的技能。适合希望将 AI 技术从概念转化为可上线产品的开发者和工程师入门参考。
Python · ★ 2,380 · 🍴 444 · 📈 226 stars today
Music Assistant is a free, opensource Media library manager that connects to your streaming services and a wide range of connected speakers. The server is the beating heart, the core of Music Assistant and must run on an always-on device like a Raspberry Pi, a NAS or an Intel NUC or alike.
中文介绍 开源媒体库管理服务端，可连接 Spotify、Tidal 等主流流媒体服务及本地音乐库，并与 Sonos、Chromecast 等多种智能音箱联动播放。作为 Music Assistant 的核心后端，适合家庭影音发烧友部署在 Home Assistant 或 NAS 上，打造统一的个人音乐管理中心。
Python · ★ 17,267 · 🍴 2,562 · 📈 361 stars today
M3U Playlist for free TV channels
中文介绍 提供免费电视频道 M3U 播放列表的资源项目，汇聚来自多个国家和地区的合法免费直播频道链接，格式兼容 VLC、Kodi、Jellyfin 等主流播放器。适合希望在不付费订阅的情况下通过软件播放器收看免费直播电视的用户。
TeX · ★ 3,054 · 🍴 655 · 📈 488 stars today
Introduction to Autonomous Robots
中文介绍 《自主机器人导论》开源教材，系统介绍自主机器人的核心理论，涵盖运动学、感知、定位、路径规划与机器学习等模块，配有图文说明和习题。适合机器人学方向的高校学生和科研人员作为入门教材，也可供工程师自学自主系统基础知识。
PowerShell · ★ 48,006 · 🍴 1,936 · 📈 114 stars today
A simple, lightweight PowerShell script that allows you to remove pre-installed apps, disable telemetry, as well as perform various other changes to declutter and customize your Windows experience. Win11Debloat works for both Windows 10 and Windows 11.
中文介绍 轻量级 PowerShell 脚本工具，用于清理 Windows 11/10 系统预装臃肿软件、禁用遥测数据上报、关闭广告推送，并支持多项系统个性化调整。操作简单，无需手动编辑注册表，适合注重隐私保护或希望获得更流畅、干净 Windows 使用体验的普通用户和 IT 管理员。
Dockerfile · ★ 20,996 · 🍴 1,062 · 📈 256 stars today
Self-Hosting Guide. Learn all about locally hosting (on premises & private web servers) and managing software applications by yourself or your organization. Including Cloud, LLMs, WireGuard, Automation, Home Assistant, and Networking.
中文介绍 全面的自托管技术指南，涵盖在本地服务器或私有云上部署和管理软件应用的方方面面，包括 LLM 本地部署、WireGuard VPN、自动化运维、容器化等主题，并附大量工具推荐。适合希望摆脱第三方 SaaS 依赖、在自有硬件上掌控数据与服务的个人用户和企业 IT 团队参考。
C# · ★ 3,673 · 🍴 158 · 📈 321 stars today
Free, open-source Windows optimization tool for performance, privacy, and simplicity.
中文介绍 免费开源的 Windows 系统优化工具，专注于性能提升、隐私保护和系统精简三大方向，提供图形界面操作，可一键调整系统设置、关闭后台服务和遥测功能。适合希望在不深入折腾注册表的情况下快速优化 Windows 系统的普通用户。
Python · ★ 6,311 · 🍴 462 · 📈 1,079 stars today
Security scanner for AI agent skills. Detect vulnerabilities, malicious patterns, and security risks.
中文介绍 NVIDIA 推出的 AI Agent 技能安全扫描工具，专门用于检测 AI Agent 技能插件中的漏洞、恶意代码模式和潜在安全风险。面向构建或部署 AI Agent 系统的开发者和安全工程师，帮助在发布前识别技能组件中的安全威胁，保障 AI Agent 生态的可信度。
Python · ★ 30,266 · 🍴 5,191 · 📈 395 stars today
Kronos: A Foundation Model for the Language of Financial Markets
中文介绍 面向金融市场语言的基础大模型 Kronos，专为理解和生成金融市场相关数据与文本而设计，旨在成为金融领域的通用预训练模型。适合量化研究员、金融科技工程师在市场预测、金融文本分析及智能投研等场景中应用和二次开发。
👍 2
Affordance reasoning, the inference of an object's action possibilities from its physical properties (e.g., shape and material), is fundamental to human physical understanding and increasingly critical for Large Language Models (LLMs). However, existing affordance benchmarks largely expose explicit
👍 3
Online group chats are social spaces with local conversational norms that are rarely stated explicitly. The ability and willingness of LLM-based agents to recognize and adapt to these norms remains mostly unexplored. We introduce LoSoNA, a benchmark for local social norm adaptation in multi-party ch
👍 0
Large reasoning models typically follow a read-then-think paradigm: they observe the complete input, reason over a static context, and then produce the answer. Yet many real-world scenarios are inherently dynamic, such as audio and video stream, where information arrives as a continuous stream and m
👍 11
Large language models (LLMs) are widely used in text-to-image (T2I) systems, but they are typically limited to text encoding, while denoising is handled by newly trained generative backbones. The emergence of representation autoencoders (RAEs) shifts the generation target toward semantically structu
👍 11
Coding agents powered by large language models have demonstrated strong performance on software engineering tasks. Yet most agents consume repositories almost entirely as text, which differs from how human developers use visual structure such as folder hierarchies and dependency relationships to ori
👍 33
AI agent performance depends critically on the runtime harness, comprising the prompts, tools, memory, and control flow that mediate how a model observes, reasons, and acts. Yet today's harnesses remain largely hand-crafted and static: each new model or task still demands bespoke scaffolding, and th
👍 4
Building trustworthy medical multimodal large language models (MLLMs) is critical for reliable clinical decision support. Existing medical hallucination benchmarks mainly focus on data collection, but often ignore where hallucinations originate within the reasoning process. We find that hallucinatio
👍 42
Large Language Models (LLMs) are undergoing a fundamental transformation from conversational generators into integrated AI systems capable of reasoning, action, memory, and self-improvement. We conceptualize this transition as a shift from Chatbot to Digital Colleague: from conversational answers to
👍 32
Retrieval-augmented generation is moving beyond text into long, egocentric video, where systems must select query-relevant chunks across multiple modalities and temporal granularities. Yet progress in VideoRAG is limited by two gaps: existing benchmarks allow queries to be answered without the video
👍 91
Cloning camera motion from reference videos is an important task in video generation, as videos provide intuitive and precise control. Existing methods either directly use parametric representations that fail to handle multi-shot generation or synthesize cross-paired data, which suffer from data sca
👍 5
In Low-Rank Adaptation (LoRA), the scaling factor α is often treated as a mere complement to the learning rate, yet its role in optimization remains poorly understood. In this paper, we reveal that the scaling factor α and the learning rate function differently, with α emerging as the dominant drive
👍 63
Recent advances in agentic Reinforcement Learning (RL) have substantially improved the multi-turn tool-use capabilities of large language model agents. However, most existing methods assign credit over coarse heuristic units, such as tool-call boundaries or fixed workflows, making it difficult to id
👍 0
Token-level hallucination detectors are evaluated as classifiers, by AUC over all tokens, yet a streaming monitor is judged by its reaction time: the number of tokens that pass between the onset of a hallucination and the alarm. We formulate hallucination onset detection as a quickest change detecti
👍 7
Modern Lean theorem provers achieve strong performance only with substantial training and inference compute, driven in part by scarce verified proof data and the long reasoning traces of formal proof search, making both supervised fine-tuning (SFT) and sampling expensive. We introduce Pythagoras-Pro
👍 15
Large language models (LLMs) now reach expert-level scores on medical licensing exams, encouraging the assumption that high scores imply safe medical judgment while patients increasingly use them for health advice. We show this assumption is fragile: when misleading context is injected into question
👍 2
AI agents are increasingly being developed to accelerate scientific discovery, yet their practical capabilities in real research settings remain poorly understood. Existing benchmarks for AI agents rarely capture the complexity, heterogeneity, and extended reasoning required by scientific work, wher
👍 15
Users rely on execution traces to observe agent behavior, diagnose failures, and ensure accountability. These traces contain rich procedural detail, including tool invocations, intermediate decisions, and error-recovery logic. Yet this detail can expose private procedural skills, allowing downstream
👍 37
The recent success of agent swarms has shifted the paradigm of large language model (LLM)-based agents from single-agent workflows to multi-agent systems, highlighting the importance of agent orchestration for task decomposition and collaboration. However, existing orchestration frameworks are limit
👍 23
Over the last decade, building human-level artificial general intelligence has moved from far-fetched speculation to being a concrete next-decade target for many of the largest AI organisations. Achieving this goal would have profound and far-reaching impacts on human society, which raises many comp
👍 6
As AI systems built from multiple language-model agents become more common, they are increasingly used to make decisions together: discussing, negotiating, and acting on shared tasks. While individual agents may appear well-aligned when tested on their own, problems can arise from how they interact
👍 2
Multimodal large language models can write code to produce complex programs as well as use programs to do 3D modeling, which opens up a new avenue for 3D generation powered by their priors, world knowledge and reasoning. Yet existing benchmarks rarely evaluate 3D modeling through code. Such modeling
👍 9
Embodied world models have emerged as a pivotal paradigm for visual robotic decision-making and interactive environment simulation. However, conventional embodied frameworks rely on low-dimensional structured action vectors (e.g., joint angles and end-effector poses), which suffer from limited expre
👍 1
Multimodal Foundation Models (MFMs) have made substantial progress, yet remain fragile in spatial reasoning over the physical world. A key bottleneck lies in their inability to transform local egocentric observations into a global allocentric spatial representation. To address this, we propose AlloS
👍 55
Despite recent progress, LLM agents still struggle with reasoning over long interaction histories. While current memory-augmented agents rely on a static retrieve-then-reason paradigm, this rigid pipeline design prevents them from dynamically adapting memory access to intermediate evidence discovere
👍 15
Large language models (LLMs) perform inference by following a fixed depth and order, non-recurrent execution of all layers. We reveal the wide existence of training-free, flexible, dynamic program-of-layers (PoLar), where pretrained layers can be packed as modules and then skipped or looped to form
👍 0
With PRECISE, we extended Prediction-Powered Inference to produce bias-corrected estimates of ranking evaluation metrics by combining a small human-labeled set with a large LLM-judged set. PPI is provably unbiased regardless of the LLM judge's error profile. We make it applicable to hierarchical met
👍 19
We identify a new dimension for enhancing rollout diversity in Group Relative Policy Optimization (GRPO) for LLMs. While GRPO relies on diverse rollouts, prevailing strategies primarily increase diversity by injecting more token-level randomness, which may introduce step-wise noise and lead to incoh
👍 2
Studies of human reasoning have shown that people are typically stronger at evaluating reasoning than producing it from scratch. In contrast, large reasoning models (LRMs) are trained to excel at producing long chains of reasoning to solve complex problems. How then do LRMs perform at evaluating rea
👍 0
We present a benchmark for evaluating AI models and agents on real-world formal software verification tasks. We first scrape 11,039 property-based tests (PBTs) from real-world Python repositories, then automatically translate 2,772 of them (25%) into 9,415 Lean 4 specifications with sorry placeholde
👍 0
Interactive driving exposes a failure mode that is easy to miss in rule-aware autonomous-driving stacks: a hard-rule margin can be negative for an ego candidate even though a small lawful accommodation by a non-priority agent would restore feasibility. Existing rulebooks, shields, and reachability f
@satyanadella · 5.9M 粉丝 · 1.9M 阅 · 3.5K 赞 · 541 转
I’ve been thinking a lot about the future of the firm in an AI-driven economy. This transition is different than any previous platform shift. In the past, we used digital systems to enhance human
中文介绍 微软CEO纳德拉深度思考AI经济下企业形态的变革。他认为本轮AI浪潮与以往平台转型根本不同——数字系统过去用于辅助人类，而现在AI正在重塑组织结构与价值创造方式。核心论点是：没有生态系统支撑的前沿技术终将不稳定，暗示微软将押注构建AI时代的平台生态而非单点能力。
@BradGroux · 5.9K 粉丝 · 714.6K 阅 · 1.0K 赞 · 638 转
Most people still use coding agents like fancy autocomplete or a one-shot chat box. That leaves a lot of value on the table. The better pattern is to treat Codex like a durable operating loop:
中文介绍 分享将 Codex 当作「持久运行循环」而非一次性聊天工具的使用范式。作者认为大多数人把 Codex 当高级自动补全，严重低估其潜力。正确姿势是把它嵌入可持续运转的操作闭环，而非单次问答——属于 Codex 工作流进阶方法论分享。
@Fintech03 · 34.9K 粉丝 · 438.8K 阅 · 501 赞 · 137 转
The critique that Indian IT services (the classic TCS, Infosys, Wipro, HCL cohort) are failing India in the current generative AI race is an incredibly popular talking point. It is easy to look at the
中文介绍 反驳「印度IT巨头（TCS、Infosys、Wipro、HCL）在AI竞赛中失败」的流行观点。作者认为这一批评混淆了定位——印度IT服务业从未以构建基础模型为目标，拿 ChatGPT 类创新来衡量它们本质上是错误参照。是一篇为印度IT行业正名的分析性推文。
@AnthonyNAguirre · 4.2K 粉丝 · 232.1K 阅 · 527 赞 · 38 转
I've been getting really bothered lately by something subtle but quite dangerous that I believe is currently going on and promises to get worse: Large numbers of very smart, capable, and important
中文介绍 物理学家 Aguirre 警示一种隐蔽但危险的趋势：大量聪明、有影响力的人正在经历某种「精微洗脑」，且情况将持续恶化。具体所指未在预览中展开，但语气严肃，指向AI或科技领域的认知偏差与群体思维问题，属观点性预警帖。
@omarsar0 · 307.3K 粉丝 · 81.2K 阅 · 518 赞 · 66 转
Autonomous coding is moving from better prompting to better control systems. The important shift is that engineers are learning how to wrap agents in goals, evaluators, loops, and artifacts that let
中文介绍 梳理自主编程 Agent 的演进方向：从「更好的提示词」转向「更完善的控制系统」。工程师们正学会将 Agent 包裹在目标设定、评估器、运行循环与产物管理等机制中，使其真正实现长时间自主运行。属于编程 Agent 工程实践的趋势分析。
@AndrewCurran_ · 53.9K 粉丝 · 62.8K 阅 · 569 赞 · 61 转
If you used Fable while it was available, you know it is special in ways that will not show up on benchmarks. I post benchmarks all the time because they matter to many people, but for a long time
中文介绍 感叹 AI 工具 Fable 已下线，作者认为它具备无法被 benchmark 衡量的独特价值。借此引出一个更深层观点：长期以来 benchmark 主导了公众对模型好坏的判断，但真正重要的用户体验维度往往被数字掩盖。语气带有惋惜与反思色彩。
@addyosmani · 400.1K 粉丝 · 52.9K 阅 · 522 赞 · 41 转
Coding agents are extraordinarily good now and getting better fast. The interesting consequence is that the hard part of engineering moved from writing code to deciding whether to trust it, which
中文介绍 Google工程师 Addy Osmani 探讨 Agent 编码能力飞速提升带来的工程范式转变：难点已从「写代码」转移到「判断是否信任代码」。Code Review 进入 Agentic 时代，工程师的核心价值在于甄别与决策，而非生产。观点简洁有力，切中当前行业痛点。
@unicity_labs · 125.6K 粉丝 · 47.6K 阅 · 534 赞 · 659 转
Programmatic agent usage moves onto a separate metered credit today. Here is what changes, and the controls we are building into AstridOS to manage it. What changed today Today, 15 June 2026,
中文介绍 2026年6月15日起，Claude Agent SDK 的程序化 Agent 调用正式拆分为独立计费额度。帖子解读了此次账单变更的具体影响，并介绍其产品 AstridOS 为此新增的用量管控机制。对依赖 Claude API 构建 Agent 应用的开发者有直接运营参考价值。
@zaimiri · 54.4K 粉丝 · 46.9K 阅 · 508 赞 · 48 转
Most people try to build their AI setup in one chaotic weekend. They install ten tools. Connect five APIs. Create a few automations. Add a giant system prompt. Then wonder why the whole thing feels
中文介绍 分享名为「Hermes」的个人AI系统7天搭建完整教程。作者指出大多数人一次性安装十几个工具、连接多个API、堆砌自动化流程，结果杂乱低效。Hermes 方法论主张循序渐进构建，属于个人AI工作流从零搭建的系统性指南。
@Dimillian · 51.8K 粉丝 · 38.2K 阅 · 558 赞 · 37 转
How to turn your phone into a Codex control center?? At first glance, it looks like a way to check on a Codex task from your phone. That is useful, but it misses the bigger idea. The power of Codex
中文介绍 分享如何将手机打造成 Codex 移动控制中心的进阶用法。作者认为手机端 Codex 远不止「随时查看任务进度」这么简单，真正的价值在于利用移动端实现对多个 Codex 任务的灵活调度与工程管控。属于 Codex 移动端深度使用教程。
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. When I landed in Seoul after a grueling 12-hour flight from San Francisco, I walked through an unmanned immigration checkpoint, where a machine scanned my face an
中文介绍 《麻省理工科技评论》探讨韩国对人工智能的高度热忱。文章以作者抵达首尔后经历无人值守的AI人脸识别出入境通道为引，分析韩国社会在政府政策推动和文化背景影响下，对AI技术的广泛接受与深度融合现象，展现韩国在AI应用普及方面的领先程度。
中文介绍 TLDR AI 2026年6月15日简报涵盖以下要点：Anthropic旗下AI项目Fable宣布关闭；中文大语言模型GLM-5.2发布；OpenRouter推出Fusion新功能。具体细节摘要原文未予提供。
OpenAI launches the Partner Network, investing $150M to help global partners accelerate enterprise AI adoption, deployment, and transformation.
中文介绍 OpenAI正式推出「合作伙伴网络」（Partner Network），投入1.5亿美元资金，助力全球合作伙伴加速企业级AI的采用、部署与业务转型，旨在推动人工智能在企业场景中的规模化落地应用。
We are in the strangest timeline.
中文介绍 据Latent Space报道，Anthropic旗下AI项目Fable与Mythos被官方认定风险过高，最终决定不予公开发布。这一决定引发业界广泛关注，反映出头部AI机构在安全评估层面日趋审慎的态度。
中文介绍 Allen AI在Hugging Face发布olmo-eval评估工作台，专为模型开发循环设计，旨在为研究人员提供系统化的模型评估流程与工具支持，助力开源大语言模型OLMo系列的持续迭代与性能优化。
OpenAI introduces three Academy courses that help people build practical AI skills, create repeatable workflows, and apply agents in everyday work.
中文介绍 OpenAI Academy推出三门新课程，聚焦职场AI实际应用，帮助用户掌握AI实用技能、建立可复用的工作流程，并将AI智能体融入日常工作场景，以应对新一轮AI驱动的职场变革需求。
a quiet day lets us highlight a great concept from Peter Steinberger, Boris Cherny, and Andrej Karpathy
中文介绍 Latent Space发布AINews简报，重点介绍由Peter Steinberger、Boris Cherny与Andrej Karpathy提出的「Loopcraft」概念，探讨AI开发中循环堆叠（Stacking Loops）的设计思路与方法论，为开发者提供系统性参考。
Preply uses OpenAI to launch AI-generated lesson summaries, providing personalised feedback and language learning exercises.
中文介绍 语言学习平台Preply借助OpenAI技术推出AI生成课程摘要功能，为学习者提供个性化反馈与语言练习内容，将AI与人类教师相结合，实现更具针对性的定制化语言学习体验。
中文介绍 TLDR AI 2026年6月12日简报涵盖：OpenAI宣布收购数据平台Ona；Anthropic就此前某一立场作出政策调整；小米发布MiMo代码大模型。各条目详细内容简报未进一步说明。
Google DeepMind is funding research into the potential dangers of situations where millions of different AI agents interact with each other online. According to Rohin Shah, who directs the company’s AGI safety and alignment research, the mass-market arrival of agents that can carry out tasks without
中文介绍 Google DeepMind正资助研究，探讨数百万AI智能体在网络上大规模相互交互时可能引发的潜在风险。据该公司AGI安全与对齐研究负责人Rohin Shah透露，能够自主执行任务的智能体一旦大规模普及，其群体交互行为将带来难以预测的安全隐患。
a quiet day lets us reflect on a great essay
中文介绍 Latent Space刊载投资人Sarah Guo的深度文章，探讨开放模型的发展现状、模型实验室与智能体开发商之间的竞争格局，以及AI训练中哪些能力无法通过训练习得等核心议题，为行业发展提供思考框架。
Learn how BBVA scaled ChatGPT Enterprise to 100,000 employees and partnered with OpenAI to accelerate AI-powered banking transformation worldwide.
中文介绍 西班牙国际银行BBVA与OpenAI达成深度合作，将ChatGPT Enterprise部署规模扩展至全行10万名员工，成为全球金融业大规模应用AI的标杆案例，推动银行业务向AI驱动模式全面转型。
Discover how astrophysicist Chi-kwan Chan uses Codex to build black hole simulations, helping scientists study extreme physics and test Einstein’s theory of general relativity.
中文介绍 天体物理学家Chi-kwan Chan介绍如何借助OpenAI Codex构建黑洞模拟程序，利用AI辅助编写科学计算代码，支持极端物理现象研究及爱因斯坦广义相对论验证，展示AI工具在前沿科学研究中的实际应用价值。
OpenAI plans to acquire Ona to expand Codex with secure, persistent cloud environments, enabling long-running AI agents across enterprise workflows.
中文介绍 OpenAI宣布计划收购云基础设施公司Ona，旨在为Codex扩展安全、持久化的云端运行环境，支持长周期AI智能体在企业工作流中的稳定运行，进一步强化其企业级AI智能体产品能力。
OpenAI supports the EU Code of Practice on AI content transparency, advancing provenance standards and tools to help people understand AI-generated content.
中文介绍 OpenAI宣布支持欧盟《人工智能内容透明度行为准则》，承诺推进AI生成内容的溯源标准建设，并开发相关工具帮助用户识别AI生成内容来源，以配合欧盟构建可信赖AI生态系统的政策目标。
本交易日各类资产技术面呈现明显分化格局。美股方面，SPY 单日涨 1.76%、5 日涨 2.11%，当前价 754.83 距 52 周高点仅 -0.73%，RSI14=60.4，多头排列完好；QQQ 单日更大涨 3.14%，距 52 周高点 -0.62%，两大宽基 ETF 技术面均偏强。科技股内部则出现分化：NVDA 单日涨 3.54% 且维持多头排列，而 MSFT 和 META 仍处于空头排列，MACD 柱均为深度负值，短线反弹尚未扭转中期趋势。VIX 恐慌指数单日大跌 8.37%、5 日跌 14.38%，当前报 16.2，低于 SMA20（17.45）和 SMA50（18.09），恐慌情绪明显缓和，对美股形成技术面支撑背景。加密货币方面，恐慌贪婪指数仅为 23（极度恐慌），加密总市值 2.35T 美元，BTC 主导率 56.5%。BTC 当前价 66,261，仍处空头排列（SMA50=73,672 / SMA200=77,521），但 MACD 已形成金叉，ETH、SOL 同样出现 MACD 金叉信号，显示短线动量有所修复，但结构性下行格局未改。中概股整体技术面偏弱，BABA RSI14 降至 29.4 触发超卖，多只标的空头排列；腾讯（0700.HK）5 日涨 2.96% 略有回稳，但 MACD 仍低于信号线。商品与外汇方面，WTI 原油 5 日重挫 10.74%，跌破多条均线；黄金期货当日涨 2.93%，但 MACD 仍处负值区间；美元指数 DXY 维持多头排列，RSI14=56.5，10 年期美债收益率（4.47%）5 日小幅回落 1.82%，美元/人民币汇率触及接近 52 周低位。
QQQ 单日涨幅达 3.14%，5 日涨幅 3.9%，当前价 744 元距 52 周高点仅 -0.62%，处于多头排列（SMA20=723.25 / SMA50=684.99 / SMA200=626.23 均线依序向上）。RSI14 为 62.5，尚未进入超买区域，但 MACD（9.73）仍低于信号线（12.83），MACD 柱处于绿柱收窄阶段，短期动量尚未完全恢复，需关注 MACD 是否形成金叉以确认上行延续。
BABA 当前价 112.55，5 日跌幅 -6.26%，距 52 周高点 -41.58%，三条均线呈空头排列（SMA20=124.81 / SMA50=130.14 / SMA200=149.53 均高于现价）。RSI14 低至 29.4，已触发「RSI 超卖」信号，但 MACD（-5.24）仍在信号线（-3.73）下方运行，空头趋势未见明确反转迹象。超卖状态下存在技术性反弹可能，但整体指标仍偏弱。
SOL 单日涨幅 7.34%、5 日涨幅 13.8%，短线动量强劲，但当前价 73.93 距 52 周高点仍达 -70.8%，SMA20=72.54 / SMA50=81.46 / SMA200=99.68 呈空头排列。MACD（-4.55）已上穿信号线（-4.93）触发金叉，RSI14 为 44.4，尚在中性区间。技术面呈现短线反弹与中长期空头并存的分歧格局，方向性尚不明确。
META 当前价 593.48，虽单日涨 4.67%，但 5 日仅涨 1.38%，距 52 周高点 -25.47%，三条均线呈空头排列（SMA20=603.17 / SMA50=622.21 / SMA200=657.32 均高于现价）。MACD 为 -11.67，信号线 -8.71，MACD 柱深度负值，下行动量仍较显著。RSI14=46.3，尚未超卖，趋势标记为「bearish」。单日反弹幅度与整体空头排列格局形成背离，指标偏弱。
WTI 原油当前价 81.49，单日跌幅 -3.99%、5 日跌幅 -10.74%，价格已大幅跌破 SMA20（92.78）和 SMA50（96.14），MACD（-3.20）低于信号线（-2.13），绿柱延伸。RSI14=34.7，接近超卖边界但尚未触发，下行趋势明显。SMA200=73.5 为下方较重要支撑参考位，短期技术面持续偏弱。
VIX 恐慌指数
10Y 美债收益率 (%)
美元指数 DXY
S&P 500 ETF
Nasdaq 100 ETF
Apple
Microsoft
Nvidia
Alphabet
Tesla
Meta
Bitcoin
Ethereum
Solana
阿里巴巴 (BABA)
拼多多 (PDD)
京东 (JD)
腾讯控股 (0700.HK)
黄金期货
WTI 原油期货
美元 / 人民币
本报告所有内容均基于公开行情数据计算的技术指标，仅供技术指标解读参考，不构成任何投资建议或买卖推荐。过去走势不代表未来表现，市场存在不确定性，投资者应自行评估风险。
President Trump and Iran’s lead negotiator have signed an agreement that was expected to open the Strait of Hormuz. But the specific terms of the deal have not been released, and Israel’s leader appears not to be fully on board.
中文摘要 美国总统特朗普与伊朗首席谈判代表签署初步协议，预计将重新开放霍尔木兹海峡。但协议具体条款尚未对外公布，以色列总理内塔尼亚胡似乎未完全认可该协议。
Democrats demanded an immediate briefing and even Republicans conceded they had no information on an agreement the administration has declined to release.
中文摘要 美国国会参议员对美伊协议内容知之甚少，特朗普政府拒绝公布协议细节。民主党议员要求立即举行情报简报会，共和党议员也承认对协议内容毫不知情，多数参议员暂未对协议表示支持。
“The struggle has not ended,” a defiant Prime Minister Benjamin Netanyahu told Israelis.
中文摘要 以色列总理内塔尼亚胡表示，尽管美伊已签署协议，以色列军队将继续驻留黎巴嫩，强调「斗争尚未结束」。内塔尼亚胡态度强硬，拒绝立即撤军。
French president Emmanuel Macron met with Trump and congratulated him for signing the new Iran agreement.
中文摘要 特朗普抵达法国出席七国集团峰会，并大力宣传与伊朗签署的和平协议。法国总统马克龙与特朗普会面，对美伊协议表示祝贺。
But overall interest in news has increased, particularly among women and young people, 2026 Digital News Report finds Follow our Australia news live blog for latest updates Sign up for Guardian Australia’s free weekly media newsletter here The majority of Australian adults under 25 have never used n
中文摘要 据2026年澳大利亚数字新闻报告，大多数25岁以下澳大利亚人从未通过报纸或广播获取新闻。但整体新闻关注度有所上升，女性和年轻群体的兴趣尤为明显。
Prime Minister Keir Starmer said his government planned to bar children under 16 from social media, following similar efforts in Australia and elsewhere.
中文摘要 英国首相基尔·斯塔默宣布，政府计划禁止16岁以下未成年人使用社交媒体，继澳大利亚等国之后推出类似限制措施。
Welsh singer, best known for 1983 hit Total Eclipse of the Heart, had emergency intestinal surgery in May Welsh pop star Bonnie Tyler is no longer in a coma but remains “very unwell” in intensive care at a hospital near her home in Faro, Portugal. The 75-year-old singer received emergency intestinal
中文摘要 威尔士歌手邦妮·泰勒已从昏迷中苏醒，但仍在葡萄牙法鲁附近医院重症监护室接受治疗，状态「非常虚弱」。这位75岁歌手因1983年金曲《Total Eclipse of the Heart》享誉国际，今年5月曾接受紧急肠道手术。
The short-term truce hasn’t been made public, and the long-term peace deal it is supposed to yield doesn’t exist yet, so there is much still unknown about any agreement between the adversaries.
中文摘要 美伊短期停火协议内容尚未公开，预期由此达成的长期和平协议也尚不存在，外界对两国协议的诸多细节仍一无所知。协议的实际效力与执行机制存在大量未解疑问。
Charging a toll is illegal under international law, but some fees are allowed for services. It is not clear what services Iran would provide, but there were no fees charged before the war.
中文摘要 美伊协议涉及霍尔木兹海峡通行费问题，国际法禁止征收过境税，但允许对特定服务收费。目前尚不清楚伊朗将提供何种服务，战前该海峡亦未征收任何通行费用。
A journalist for Press TV has been hit by shrapnel while filming a report on Israeli attacks in southern Lebanon.
中文摘要 一名伊朗国际电视台（Press TV）记者在黎巴嫩南部拍摄以色列袭击报道时，遭以军炮击产生的弹片击中受伤。
G7 told ‘we will stand with Ukraine for as long as it takes’, with Russia’s finance networks and shadow fleet targeted Keir Starmer has vowed to “choke off” Russian revenue with further sanctions and to provide hundreds of millions of pounds worth of energy support for Ukraine, as he met world leade
中文摘要 英国首相斯塔默在七国集团峰会上承诺对俄罗斯实施新一轮制裁，目标包括俄罗斯金融网络和「影子船队」，并表示将向乌克兰提供数亿英镑核能支持，重申将与乌克兰「并肩作战直至胜利」。
After many false starts, this cease-fire plan could be different. That’s because the war really has become painful for both the U.S. and Iran.
中文摘要 分析认为此次美伊停火协议可能有别于以往多次失败尝试，原因在于持续战争已令美伊双方均付出沉重代价，双方均有较强动力维持停火。
President Trump says a deal had been reached to prolong a ceasefire with Iran and open the Strait of Hormuz, but many challenges remain.
中文摘要 特朗普宣布美伊双方已达成协议，旨在延长停火并重开霍尔木兹海峡，但协议具体条款极为有限，外界对其能否有效落实仍存诸多疑虑。
Donald Trump posts ‘Let the oil flow’ as US-Iran peace deal sparks immediate drop for Brent crude Analysis: Oil and gas unlikely to return to prewar prices for months even if Hormuz reopens Middle East crisis: live updates Global oil prices have tumbled to a three-month low and stock markets closed
中文摘要 美伊和平协议消息推动国际油价跌至三个月低点，全球股市同步创历史新高。特朗普发文称「让石油流动起来」，布伦特原油价格即时大幅下滑，但分析指即使霍尔木兹海峡重开，油气价格恢复战前水平仍需数月。
Preliminary agreement between Washington and Tehran has prompted anger in Israel, and criticism of Netanyahu Benjamin Netanyahu has hailed a historic victory over Iran and ruled out any immediate withdrawal from Lebanon, saying that Israel’s forces would remain there “for as long as necessary”. “We
中文摘要 内塔尼亚胡宣称以色列对伊朗取得历史性胜利，同时明确拒绝从黎巴嫩撤军，表示以军将「无限期」驻留当地。美伊初步协议在以色列国内引发强烈不满，内塔尼亚胡亦因此受到批评。
A single trader on Polymarket lost nearly $1 million when Cabo Verde fought Spain to a stunning draw on Monday in one of the most unlikely outcomes in recent World Cup history, powered by a 40-year-old goalkeeper who left the pitch in tears.
中文摘要 在世界杯小组赛中，佛得角与西班牙战平，堪称近年最出人意料的赛果之一。该场赛事令预测平台Polymarket上一名交易员损失近100万美元。佛得角队关键角色为一名40岁的守门员，赛后含泪离场。此次意外结果再次凸显预测市场在极端小概率事件面前的巨大风险。
A government spokesman tells the BBC that the current offer "does not do enough to protect consumers or the environment".
中文摘要 英国泰晤士水务公司国有化进程又迈进一步。英国政府发言人向BBC表示，目前的救助方案"在保护消费者和环境方面做得还不够"，正式反对现有重组协议。此前，泰晤士水务因债务危机深陷困境，政府介入的可能性持续上升。
Brazilian prosecutors rejected a cooperation agreement proposed by Daniel Vorcaro, the former banker at the center of a massive fraud investigation, according to a person familiar with the matter.
中文摘要 据知情人士透露，巴西检察官拒绝了Banco Master大规模欺诈调查核心人物、前银行家丹尼尔·沃尔卡罗提出的合作协议。沃尔卡罗此前寻求以配合调查换取司法豁免，但检方最终未予接受，案件调查仍在推进中。
Gold held a gain after US President Donald Trump said the Strait of Hormuz could reopen on Friday, likely easing an energy and inflation shock that has roiled global markets.
中文摘要 美国总统特朗普表示霍尔木兹海峡可能于本周五重新开放，相关消息提振市场情绪，缓解了此前能源价格上涨及通胀压力对全球市场的冲击。黄金在此背景下维持涨势，市场对地缘风险的定价有所调整。
Private equity firm IG4 Capital, fresh off a deal to take control of petrochemical giant Braskem SA, is now setting its sights on sugar and ethanol producer Raizen SA, another major Brazilian company in distress.
中文摘要 私募股权公司IG4 Capital在完成对石化巨头Braskem SA的控股收购后，将目标转向另一家陷入困境的巴西企业——糖与乙醇生产商Raizen SA。IG4目前提出收购Raizen债务，以寻求取得该公司多数股权。
South London football club is enjoying its most successful ever period
中文摘要 据《金融时报》报道，英超球队水晶宫的美国亿万富翁股东正在探讨出售该俱乐部的可能性。水晶宫目前正处于历史上最成功的竞技阶段，此时传出出售消息引发外界广泛关注。
Traders in Asia turned their focus to policy calls from the Reserve Bank of Australia and the Bank of Japan after a US-Iran deal to reopen the Strait of Hormuz fueled a relief rally across global financial markets on Monday.
中文摘要 美伊达成霍尔木兹海峡重开协议，推动全球金融市场周一出现宽幅反弹。亚洲交易员随后将目光转向澳大利亚储备银行（RBA）及日本银行（BOJ）即将公布的利率决议，关注两大央行政策走向对市场的后续影响。
Oil held its biggest drop in more than two weeks as traders, shippers and producers awaited details of a US-Iran deal that’s intended to pave the way for the full reopening of the Strait of Hormuz.
中文摘要 油价维持逾两周以来最大单日跌幅后继续承压，交易员、航运商及生产商均在等待美伊协议的具体细节，以确认霍尔木兹海峡全面重开的时间表和落实安排，市场方向暂时存在不确定性。
The next big growth market will align closely with domestic spending on infrastructure.
中文摘要 全球钢铁行业将下一个增长引擎的目光投向印度，但分析认为印度钢铁需求的驱动逻辑与中国有本质区别——印度增长主要依托国内基础设施建设支出，而非出口导向的大规模工业化，因此钢铁繁荣的规模和节奏将有所不同。
John Rogers, Ariel Investments founder, chair, CIO and co-CEO, says MSG stocks have not reflected the value of the Knicks or the Rangers during an interview with Romaine Bostick and Katie Greifeld on "Bloomberg The Close." (Source: Bloomberg)
On today’s Big Take podcast, we examine the major hurdles facing a proposed US-Iran peace deal, including how Israel fits into the equation.
中文摘要 彭博《Big Take》播客分析了美国、以色列与伊朗三方在和平协议谈判中的立场与分歧，重点探讨以色列在美伊和谈框架中的角色，以及达成终止战争协议所面临的主要障碍。
Jeff Simmons, Elanco Animal Health President & CEO, discusses the emerging threat of screwworm infections spreading beyond the initial control zone in Texas, with at least 12 confirmed cases including eight heads of cattle. Simmons reassured viewers that while the pest is expected to spread, it can
中文摘要 Elanco动物健康公司总裁兼CEO杰夫·西蒙斯表示，螺旋蛆虫感染已在德克萨斯州初始控制区外出现至少12例确诊病例，其中包括8头牛。他认为该害虫的管控难度大于围堵，但对外界保证局面仍可控，相关防治工作正在推进。
17 回复 · 程序员 节点
7 回复 · 程序员 节点
3 回复 · 程序员 节点
14 回复 · Apple 节点
6 回复 · Apple 节点
14 回复 · Apple 节点
7 回复 · Apple 节点
68 回复 · Apple 节点
8 回复 · Linux 节点
7 回复 · Linux 节点
戾气还是太重了 10 个帖子 - 8 位参与者 阅读完整话题
在实现手机远程控制（第三方api登录态）codex桌面版的功能开发中，几乎把坑都给踩遍了，终于，成功实现了这一功能并炼进了skills。 本该早些发布的，被四六级备考给拖住进度了 skills： 【开源】Codex桌面版部分功能(Fast Mode、插件市场、道德限制、手机远程控制和 Computer Use 等)被限制？一个skills解决！【6.15重磅更新】 使用教程： 先仔细阅读skills使用说明，然后让agent进行修复。 修复完成后需要在代理的环境下，打开设置->连接->添加，等待弹出二维码，用登录了chatgpt的手机扫描二维码（账号需要开启“多重身份验证”和“为 Codex
28 个帖子 - 23 位参与者 阅读完整话题
好几天了这是第一次蹬上any 佬们，这是何意味？ *看到佬们几分钟这么多回复，声明一下不是质疑，就是遇到了图一乐昂。按楼下说的很可能是第一次say hi的时候选了肥波5的原因，导致第二次问你是什么模型的时候，它可能附带了什么提示词了 咱也不懂，应该就素酱紫。 50 个帖子 - 44 位参与者 阅读完整话题
本帖使用社区开源推广，符合推广要求。我申明并遵循社区要求的以下内容： 我的帖子已经打上 开源推广 标签： 是 我的开源项目完整开源，无未开源部分： 是 我的开源项目已链接认可 LINUX DO 社区： 是 我帖子内的项目介绍，AI 生成、润色内容部分已截图发出： 是 以上选择我承诺是永久有效的，接受社区和佬友监督： 是 佬友们，接着https://linux.do/t/topic/2394586 讨论 为了高效的科研，尽量发论文毕业，我从 openai 刚出来就在使用各种 AI 模型，并且从今天年初使用 openclaw 的时候就在研究迭代各种 skill。本着开源的态度，我把自己的这套 “自
今年1月份投稿的Elsevier论文，今天终于中了，特作纪念。 说来惭愧，这篇论文还是24年7月份开始构思和做实验的。因为实验结果一直不理想，一直拖到25年才终于定下初稿。一开始投的1区，后来不出所料被拒了。继续修改投了现在这个2区（具体的名字就不说了）经过二审后终于录用。直到现在打开doi号看到自己的网页版和可以下载的pdf版，依然还有一种不真实的感觉。这是我的第一篇论文，我还是很用心的，像是自己的孩子一样。至于论文是什么，就不给佬们献丑了。 最后祝L站的佬友们，也能论文顺利，一投就中 28 个帖子 - 14 位参与者 阅读完整话题
本来因为我号池PLUS号不多了，已经关闭了GPT渠道 做公益站真的很矛盾，每次看到号池见底 这个是昨天新搭建的服务器渠道 18 个帖子 - 18 位参与者 阅读完整话题
事先声明：我并非有意带节奏，但是接连问了两个管理都没有人理我（可能是比较忙），这事对我真的很重要，迫于无奈，我只好发帖求助 事情是这样的，我从去年开始追求连续365天上线，可是Discourse论坛系统出了问题，我在阅读时间为0的这一天很明确的上线了，而且不是上线了事，是浏览了不少帖子 我已经连着上了10个月了，这是最后两个月，我不想因为这种bug断掉，只好发帖求助 49 个帖子 - 30 位参与者 阅读完整话题
常说 因为专注 所以专业 我们成立这家ccmax20.cc中转站的目的就是想给各位佬友提供优质的服务 我们经过多月自研号池 现已稳定 不混池 是纯血cc就是纯血 号池可查 正规渠道 非速刷 倍率1 充值1RMB到账1U 倍率真实可查 拉账单 部分地区无法进入 开关魔法 飞行模式后尝试 给各位佬友安排了10刀测试余额 留下你的用户名 给你专属留言兑换码 企业对接 可开票 拉专属对接服务群 服务不中断 售后有保障 量大高并发拿直连 网站：ccmax20.cc 轻松 好记 支持LinuxDo登录 企业对接享专属倍率 对接微信：baby89ii QQ群：868616225 备注来自(linuxdo)
先攒着再说，等能用的时候狠狠狠的蹭一下~ 新佬们，我不是这个公益站站长，我跟你们一样是用户，我只是提醒大家一下今天新增了签到。 51 个帖子 - 38 位参与者 阅读完整话题