每日简报

2026-09-18

← 历史归档

alibaba/open-code-review

Go · ★ 34,714 · 🍴 2,470 · 📈 3,286 stars today

Fast, efficient, battle-tested at Alibaba's scale. Hybrid architecture code review tool: deterministic pipelines + LLM Agent, precise line-level comments, built-in multi-language ruleset (NPE, thread-safety, XSS, SQL injection), OpenAI & Anthropic compatible.

中文介绍 阿里开源的大规模代码审查工具,采用混合架构:确定性流水线负责规则匹配,LLM Agent 负责语义分析。可精确到代码行级别给出评论,内置多语言规则集,覆盖空指针、线程安全等常见缺陷,适合在大型研发团队中做 CI 集成和自动化评审。

cloudflare/security-audit-skill

JavaScript · ★ 10,621 · 🍴 566 · 📈 3,607 stars today

A coding-agent skill for multi-phase security audits with independently verified, machine-readable findings

中文介绍 Cloudflare 推出的编程 Agent 技能包,用于执行多阶段安全审计。它把审计流程拆成若干阶段,产出机器可读、且可独立验证的发现结果,适合安全团队将自动化审计接入 AI 编码助手工作流。

addyosmani/agent-skills

JavaScript · ★ 95,842 · 🍴 10,145 · 📈 680 stars today

Production-grade engineering skills for AI coding agents.

中文介绍 面向 AI 编程 Agent 的生产级工程技能集合,由 Addy Osmani 维护。提供可直接复用的技能定义,让 Agent 在真实项目中遵循规范化的工程实践,适合使用 Claude Code、Cursor 等工具的开发者扩展 Agent 能力。

Tencent/BrowserSkill

TypeScript · ★ 4,123 · 🍴 292 · 📈 1,302 stars today

Let AI agents use your real, logged-in browser without interrupting your work. CLI + extension for browser automation across any shell-capable AI agent.

中文介绍 腾讯开源的浏览器操作技能,让 AI Agent 直接调用你已登录的真实浏览器,且不打断你的正常使用。提供 CLI 与浏览器扩展,兼容任意可执行 shell 的 AI Agent,适用于需要登录态的网页自动化场景。

alphaXiv/OpenResearch

Rust · ★ 4,946 · 🍴 306 · 📈 939 stars today

Turn your coding agents into research agents

中文介绍 把现有的编程 Agent 转变成科研 Agent 的工具,来自 alphaXiv。目标是在论文检索、阅读与实验流程中复用编码 Agent 的能力,适合研究者用 AI 辅助文献调研和复现实验,降低科研自动化的接入门槛。

anthropics/claude-code

TypeScript · ★ 145,866 · 🍴 23,635 · 📈 538 stars today

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.

中文介绍 Anthropic 官方的终端内智能编程工具,能理解整个代码库,执行日常编码任务、解释复杂代码并处理 git 工作流。开发者可在命令行中与它协作开发,把重复性工作交给 Agent,从而加快编码节奏。

NationalSecurityAgency/ghidra

Java · ★ 78,476 · 🍴 8,684 · 📈 912 stars today

Ghidra is a software reverse engineering (SRE) framework

中文介绍 美国国家安全局开源的软件逆向工程框架,提供反汇编、反编译、图形化分析与脚本扩展能力,支持多种处理器架构和可执行格式。安全研究员和恶意软件分析师常用它做二进制逆向、漏洞挖掘和固件审计。

anthropics/knowledge-work-plugins

Python · ★ 24,557 · 🍴 2,941 · 📈 287 stars today

Open source repository of plugins primarily intended for knowledge workers to use in Claude Cowork

中文介绍 Anthropic 开源的插件仓库,主要面向知识工作者,用于在 Claude Cowork 中扩展能力。通过插件形式把文档处理、信息整理等任务接入 AI 协作环境,适合非工程背景的知识型团队提升日常工作效率。

Tencent/WeKnora

Go · ★ 26,196 · 🍴 3,550 · 📈 1,125 stars today

Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.

中文介绍 腾讯开源的 LLM 知识平台,把原始文档转化为可查询的 RAG、自主推理 Agent 以及可自我维护的 Wiki。适合企业搭建私有知识库,用于文档问答、知识检索和自动整理,降低大模型落地知识管理的门槛。

abue-ammar/tinycast

Swift · ★ 6,143 · 🍴 286 · 📈 739 stars today

Tinycast — a tiny, fully native macOS launcher, hotkeys, and clipboard history.

中文介绍 一款轻量的原生 macOS 启动器,同时整合快捷键与剪贴板历史功能。完全原生实现,体量小、启动快,适合希望用键盘高效唤起应用、检索剪贴板记录的 Mac 用户替代系统自带 Spotlight。

cilium/cilium

Go · ★ 25,264 · 🍴 4,073 · 📈 111 stars today

eBPF-based Networking, Security, and Observability

中文介绍 基于 eBPF 的云原生网络、安全与可观测性方案,为 Kubernetes 提供高性能网络转发、细粒度网络策略和流量可视化。广泛用于容器集群的网络插件与零信任安全加固,是 CNCF 毕业项目之一。

jamiepine/voicebox

TypeScript · ★ 54,845 · 🍴 6,825 · 📈 667 stars today

The open-source AI voice studio. Clone, dictate, create.

中文介绍 开源 AI 语音工作室,支持声音克隆、听写与语音生成。把多种语音能力整合到一个界面中,适合内容创作者、播客制作者以及需要配音或转写素材的开发者,在本地或自托管环境中完成音频处理。

affaan-m/ECC

JavaScript · ★ 261,164 · 🍴 39,095 · 📈 1,171 stars today

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

中文介绍 面向 Agent 框架的性能优化系统,提供技能、直觉、记忆、安全与研究优先的开发范式,兼容 Claude Code、Codex、Opencode、Cursor 等工具。适合希望提升 AI 编程助手执行效率与长期记忆能力的开发者。

roboflow/supervision

Python · ★ 50,802 · 🍴 4,828 · 📈 329 stars today

We write your reusable computer vision tools. 💜

中文介绍 Roboflow 维护的计算机视觉工具库,提供可复用的检测、分割、跟踪与可视化组件。可与主流推理模型和框架配合,帮助开发者快速搭建视频分析与图像处理流水线,常用于目标检测项目的后处理环节。

JustVugg/colibri

C · ★ 35,738 · 🍴 3,752 · 📈 873 stars today

Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦

中文介绍 用纯 C 编写、零依赖的推理引擎,让前沿 MoE 大模型能在自有硬件上运行,专家权重按需从磁盘流式加载。核心思路是小引擎驱动大模型,适合显存有限但希望本地部署混合专家模型的开发者。

TencentCloud/Octop

Python · ★ 3,432 · 🍴 354 · 📈 367 stars today

A smarter, self-hosted AI assistant — multi-user, multi-agent.

中文介绍 腾讯云推出的自托管 AI 助手,支持多用户与多 Agent 协作。相比单点式助手,更强调团队共享与权限隔离,适合企业内部自建 AI 助理平台,在私有环境中统一管理多个 Agent 和用户会话。

ever-co/ever-gauzy

TypeScript · ★ 7,541 · 🍴 1,120 · 📈 470 stars today

Ever® Gauzy™ - Open Business Management Platform (ERP/CRM/HRM/ATS/PM) - https://gauzy.co

中文介绍 开源的企业经营管理平台,覆盖 ERP、CRM、HRM、招聘(ATS)与项目管理等模块。适合中小企业或团队以自托管方式搭建一体化业务系统,替代多套彼此割裂的商业软件,降低采购与集成成本。

cline/cline

TypeScript · ★ 68,552 · 🍴 7,415 · 📈 380 stars today

Autonomous coding agent as an SDK, IDE extension, or CLI assistant.

中文介绍 自主编码 Agent,提供 SDK、IDE 扩展和命令行助手三种形态。可理解项目上下文并自主执行多步编码任务,适合开发者按需选择集成方式,把 AI 代理嵌入编辑器和自动化流程中提升开发效率。

coder/coder

Go · ★ 14,829 · 🍴 1,486 · 📈 145 stars today

Secure environments for developers and their agents

中文介绍 为开发者和其 AI Agent 提供安全隔离的开发环境,基于 Terraform 定义环境模板,在云端或自托管基础设施上按需创建。适合远程团队统一开发环境、加强代码与数据隔离,并支撑 Agent 的大规模运行。

n8n-io/n8n

TypeScript · ★ 204,980 · 🍴 60,745 · 📈 281 stars today

Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.

中文介绍 采用 fair-code 模式的工作流自动化平台,内置原生 AI 能力,可将可视化编排与自定义代码结合。支持自托管或云端部署,提供 400 多个集成,适合团队搭建自动化流程并接入大模型处理节点。

Continual Learning Mechanisms Compose for Long-Horizon Memorization

👍 295

Language models may need to internalize information that arrives over time and retain it through many subsequent updates. To study this challenge, we introduce long-horizon memorization, a setting in which a model learns 100 query-answer tasks through continual supervised fine-tuning without retaini

中文介绍 研究语言模型在持续监督微调中长时间保留新增信息的能力,提出「长时程记忆」设定:模型不借助回放数据依次学习100个问答任务。工作考察多种持续学习机制的组合效果,验证机制叠加可缓解灾难性遗忘,提升长时程记忆保持率。

AI for Games in the Foundation Model Era

👍 113

Foundation models, alongside advances in learned game-world models, are reshaping AI across the game lifecycle. Beyond playing games, recent systems model players and game dynamics, support design and development, adapt player-facing experiences at runtime, and evaluate resulting artifacts. Yet thes

中文介绍 综述基础模型与可学习游戏世界模型如何重塑游戏全生命周期:不仅玩游戏,还能建模玩家与游戏动态、辅助设计开发、运行时自适应调整玩家体验并评估产物。文章梳理这些系统的能力与局限,指出统一评测与落地仍待解决的问题。

StepAudio 3 Realtime Technical Report

👍 102

Realtime spoken interaction demands deep reasoning, prompt responses, and fluid turn-taking. We present StepAudio 3 Realtime, an audio-language foundation model organized around a continuous listen-converse-think-act loop. Deep Perception captures rich acoustic cues to interpret user intent, while S

中文介绍 提出实时语音交互的音频-语言基础模型 StepAudio 3 Realtime,围绕「听-说-思-行」连续循环组织。深度感知模块捕捉声学线索以理解用户意图,配合流式推理与流畅轮次切换,降低响应延迟,实现低延迟的实时对话与任务执行。

The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement

👍 88

Recursive self-improvement (RSI) enables AI systems to turn experience and feedback into persistent changes that improve both their capabilities and the process of future improvement. We first use the Headroom-Closed Index (HCI) to reveal the problems of existing LLMs, then introduce the RSI concept

中文介绍 探讨递归自我改进(RSI),即AI将经验与反馈转化为持久改变,同时提升能力与改进过程本身。作者提出 Headroom-Closed Index(HCI)指标揭示现有LLM的改进空间问题,并给出RSI的概念框架与实现路径,指向真正自主演化的智能系统。

StepAudio 3 Music Technical Report

👍 76

We introduce StepAudio 3 Music, a large-scale, long-form music generation model that supports explicit musical planning and open-domain text-controlled generation. The StepAudio Music Tokenizer represents audio as a 50-Hz stream from a 65536-entry single codebook, using semantically informed self-su

中文介绍 提出大规模长时音乐生成模型 StepAudio 3 Music,支持显式音乐规划与开放域文本控制生成。其 Music Tokenizer 用语义引导的自监督方式,将音频表示为50Hz、单码本65536词条的流式序列,从而兼顾长时结构连贯性与文本可控性。

ScienceIDE: Turning World's Scientific Codebase into Agent Learnable Environments

👍 70

Scientific code repositories encode decades of human knowledge in executable models, methods, and tools. Yet fragmented toolchains, implicit domain conventions, and specialized correctness criteria make this knowledge difficult to convert into reliable learning experience-a challenge we call the sci

中文介绍 科学代码库蕴含数十年可执行知识,但工具链碎片化、领域约定隐式、正确性标准特殊,难以转化为可靠学习经验。ScienceIDE 将这些科学代码库改造成智能体可学习的环境,让 agent 在真实科研代码中训练与验证,缓解科学任务知识获取难题。

Generalized Agent Iteration: One Formal Framework for Iterative Policy Improvement and Recursive Self-Improvement

👍 59

When we speak of recursive self-improvement (RSI), are we speaking of a phenomenon, a mechanism, or a prospect? Towards autonomous and evolving intelligence, RSI is being claimed at many scales, while no single framework that formally describes these emerging instances exists. Its counterpart in the

中文介绍 针对递归自我改进(RSI)概念被泛用却缺乏统一形式化框架的问题,提出广义智能体迭代(GAI)框架,将迭代策略改进与RSI纳入同一形式体系。该框架可描述不同尺度上的自改进实例,为自主演化智能提供统一分析工具。

Confidence Comes from Experience: Experiential Confidence Estimation from Reasoning to Agents

👍 48

Reliable confidence estimation is increasingly central to the trustworthy deployment of language models: a calibrated estimate of the probability that an output is correct decides what to ship, what to escalate, and what to retry. Existing confidence estimators, however, share one design premise: th

中文介绍 可靠置信度估计对语言模型部署至关重要,但现有方法共享一个设计前提,限制了适用性。本文提出「经验性置信度估计」,从推理到智能体任务中利用历史经验信号校准输出正确概率,从而改善下游的发布、升级与重试决策。

ProgramDistill: From Interactive Web Apps to Verifiable Reference-Guided SWE Tasks

👍 44

Coding agents are typically evaluated with desired behavior specified through issues or instructions. In practical web development, however, agents may need to infer behavior from working software and implement it in an incomplete application. We introduce ProgramDistill, a benchmark evaluating codi

中文介绍 现有编码智能体评测多以 issue 或指令描述期望行为,而真实 Web 开发需从可用软件反推行为并在不完整应用中实现。提出 ProgramDistill,将交互式 Web 应用转化为可验证、有参考引导的 SWE 任务,为编码智能体提供更贴近实践的评测基准。

Agora: Git as Shared Memory for Collective AutoResearch

👍 39

Autonomous research loops such as AutoResearch show that one coding agent can improve a training setup unattended. Run several of them and each session starts from scratch, so more agents tend to mean more duplicated search rather than more discovery. Agora is a shared memory for such agents: resear

中文介绍 AutoResearch 等自主研究循环可让单个编码智能体无人值守地改进训练配置,但多智能体并行时各会话从头开始,造成重复搜索。Agora 以 Git 作为共享记忆,让多个研究智能体复用彼此的发现与实验结果,将重复探索转化为集体积累。

EvolveTrade: Experience-Driven Policy Refinement for Self-Evolving LLM Trading Agents

👍 30

Large language model (LLM) trading agents can combine market data, news, and executable analysis, but their behavior is often controlled by static hand-written tool-use policies that are fixed before deployment. This limits their ability to adapt how they gather evidence, invoke tools, verify signal

中文介绍 LLM 交易智能体的行为常由部署前写死的工具使用策略控制,难以适应市场变化。EvolveTrade 提出经验驱动的策略精炼方法,让智能体根据交易经验自主调整证据收集、工具调用与信号验证方式,实现自我演化的交易决策。

HypoEvolve: Genetic Algorithms Enable Multi-Agent LLMs to Discover Scientific Hypotheses

👍 25

Scientific agents contribute to hypothesis discovery by synthesizing evidence, assessing proposals, and developing new explanations. Recent systems combine scientific agents with evolutionary search through critique, comparison, and revision. However, how different forms of agent collaboration affec

中文介绍 科学假设发现常借助多智能体系统与进化搜索,通过批评、比较与修订迭代候选假设,但不同协作形式的影响尚不清楚。HypoEvolve 将遗传算法引入多智能体 LLM,系统比较协作结构对假设发现效果的作用,提升科学假设生成质量。

A Zeroth-Order Paradigm for LLM Preference Alignment

👍 22

Direct preference alignment methods are widely used to align large language models (LLMs) with human preferences because of their computational and memory efficiency. However, likelihood displacement motivates alternative ways to extract information from preference pairs with small likelihood margin

中文介绍 直接偏好对齐方法因计算与显存效率被广泛用于 LLM 对齐,但似然位移问题促使其寻找从偏好对中提取信息的新途径。本文提出零阶(zeroth-order)偏好对齐范式,在小似然间隔下利用零阶优化提取偏好信号,为对齐提供新的优化视角。

ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents

👍 20

We introduce and release ScienceBuddy, an interactive scientific research workspace that brings continually improving scientific agents into researchers' everyday workflows. ScienceBuddy supports researchers in carrying out scientific tasks while transforming their requests, feedback, and execution

中文介绍 发布交互式科研工作台 ScienceBuddy,把持续自我改进的科研智能体带入研究者日常工作流。系统在协助用户完成科研任务的同时,将请求、反馈与执行轨迹转化为改进信号,实现「递归中的递归」式自提升,使智能体随使用不断进化。

HarnessVLN: Unifying Training-Free Embodied Navigation through an Agent Harness

👍 17

Embodied navigation requires agents to interpret visual observations, accumulate spatial knowledge, and execute actions to follow instructions or locate objects. Training-based methods face generalization challenges, while training-free methods exploit multimodal large language models (MLLMs) but of

中文介绍 具身导航要求智能体理解视觉观测、累积空间知识并执行动作。训练式方法泛化受限,免训练方法虽利用多模态大模型(MLLM)却常不统一。HarnessVLN 通过统一的智能体外壳整合免训练导航流程,在指令跟随与目标定位任务上提升泛化性。

In-Context Robot Learning with VLM Agents

👍 15

Enabling robots to adapt to unfamiliar environments as readily as humans remains a moonshot goal of embodied AI. No finite collection of demonstrations can cover every task and situation a robot will encounter, making the ability to learn from context at deployment essential for generalization. Such

中文介绍 机器人要像人一样适应陌生环境,仅靠有限示范无法覆盖所有任务与情境,因此需要部署时的上下文学习能力。本文研究基于视觉语言模型(VLM)智能体的上下文机器人学习,使机器人能从少量示例与场景信息中即时推断并执行新任务。

Flattening Every Memory Peak in Long-Context Mixture-of-Experts Training

👍 10

Training a Mixture-of-Experts (MoE) model at long context or large batch size fails as soon as any one component's peak allocation exceeds device memory, so the target is every peak at once, not the average footprint. Four are left unbounded by the parallelism plans in common use, and each grows dif

中文介绍 长上下文或大 batch 下训练 MoE 模型时,任一组件峰值显存超限即失败,因此目标是同时压平所有峰值而非平均占用。作者指出现有并行方案中仍有四类峰值无界增长,并提出相应方法将其逐一压平,从而支持更长上下文与更大 batch 的 MoE 训练。

ModularRSI: Modular and Generalizable Recursive Harness Self-Improvement

👍 10

Recent work extends recursive self-improvement (RSI) to agent harnesses for long-horizon coding and terminal tasks, enabling agents to improve execution mechanisms from experience. However, generalizable harness RSI remains challenging. First, evolving harnesses on evaluation benchmarks or their sub

中文介绍 已有工作将递归自我改进(RSI)扩展到智能体外壳,用于长时程编码与终端任务,但从经验改进执行机制仍难泛化。ModularRSI 提出模块化、可泛化的外壳自改进方法,避免在评测基准或其子集上过拟合,提升跨任务迁移能力。

PANORAMA: Panoptic Grounded Captioning via Mask Proposal Selection

👍 9

Intelligent systems that act in the world require image understanding that is both comprehensive and spatially grounded. Current vision-language models (VLMs) can generate fluent and detailed image captions, but reliably associating them with image pixels remains challenging. Existing methods that c

中文介绍 智能系统需要既全面又具空间定位能力的图像理解。现有 VLM 能生成流畅详细的描述,却难以可靠关联到像素。PANORAMA 通过掩码提议选择实现全景式有据描述,将图像描述与分割掩码对应,提升 caption 的空间 grounding 精度。

Learning to Solve Hard Problems in RL for LLMs by Never Giving Up

👍 9

We demonstrate that training LLMs with RL does not improve performance equally across a dataset. RL shows large improvements on easy problems that an LLM is already good at solving, but small improvements on hard problems. We call this the Matthew Effect in RL for LLMs, after the phenomenon of cumul

中文介绍 本文发现 RL 训练 LLM 在数据集上收益不均:模型已擅长的简单题提升大,难题提升小,作者称之为 RL 中的「马太效应」。研究据此提出「永不放弃」的训练策略,引导模型在难题上持续尝试,从而改善困难任务的强化学习效果。

Its time to talk about ‘model welfare’

@mustafasuleyman · 1.1M 粉丝 · 204.3K 阅 · 502 赞 · 86 转

AIs are not conscious. They do not feel, experience, or suffer. However, there’s a growing movement in support of model welfare; the idea that we might soon owe them a duty of care. I think this

中文介绍 微软 AI 负责人 Mustafa Suleyman 发文讨论「模型福祉」(model welfare):他明确表态 AI 并无意识、不会感受或受苦,但承认要求对模型承担照顾义务的呼声正在壮大,并阐述自己对此议题的立场与判断。

THE ARCHITECTURE OF AI-NATIVE WEB3.0

@LingoAITech · 411.4K 粉丝 · 156.5K 阅 · 524 赞 · 227 转

From the Semantic Web to the Human-Centric Agentic Economy Data3.0 + Finance3.0: Sovereign Data, Personal AI, Open Agents and Programmable Value Abstract Artificial intelligence is changing the role

中文介绍 一篇论述「AI 原生 Web3.0 架构」的长文摘要:从语义网演进到以人为本的智能体经济,提出 Data3.0 与 Finance3.0 的组合,核心概念是主权数据、个人 AI、开放智能体与可编程价值。

Adopting the software factory model: crawl, walk, run

@zachlloydtweets · 20.4K 粉丝 · 123.1K 阅 · 500 赞 · 47 转

The software factory approach (a closed agentic loop that runs in the cloud) is growing in popularity, but it can be daunting to adopt. In this post I’ll go through “crawl, walk, run” steps for making

中文介绍 Zach Lloyd 分享软件工厂(software factory)模式的落地路径:把云端闭环智能体循环按「爬、走、跑」三阶段推进,帮助团队循序渐进地采用这套开发范式,属于方法论型实操指南。

How we built Hermes to support our entire team

@JacquelineSYC19 · 1.0K 粉丝 · 72.4K 阅 · 501 赞 · 47 转

At the time of writing, Artie is a team of 17. Every one of us works alongside Hermes, an open-source AI agent harness from @NousResearch. It runs on a single physical machine in Germany, has a

中文介绍 Artie 团队(17 人)分享如何全员使用开源 AI 智能体框架 Hermes(出自 NousResearch)。它部署在德国单台物理机上,成为每个人的日常协作伙伴,属于小团队自建智能体基建的实操案例。

The Anatomy of Harness Engineering: How to Evaluate, Iterate, and Guard AI Coding Agents

@GoogleCloudTech · 1.3M 粉丝 · 43.8K 阅 · 505 赞 · 60 转

When developers first work on harness engineering for agentic coding systems, they often fall into the same trap: they run common end-to-end benchmarks like Terminal-Bench and DeepSWE, watch a

中文介绍 Google Cloud 技术账号解析「harness engineering」(智能体脚手架工程):开发者常犯的错是只跑 Terminal-Bench、DeepSWE 等端到端基准测试,文章讲述如何评估、迭代并防护 AI 编码智能体。

Introducing the DeepMind Institute

@ShaneLegg · 82.3K 粉丝 · 40.6K 阅 · 585 赞 · 86 转

We are on the cusp of a profound transformation. Today’s AI systems have impressive capabilities and the rapid pace of innovation suggests we’re now approaching artificial general intelligence (AGI),

中文介绍 DeepMind 联合创始人 Shane Legg 宣布成立 DeepMind Institute。他表示当前 AI 能力已相当可观,创新节奏显示人类正逼近 AGI,机构将围绕这一转折期展开工作,属于官方发布性质。

Build real-time voice applications with Gemini 3.8 Live and 3.5 Transcribe

@GoogleAIStudio · 203.4K 粉丝 · 35.5K 阅 · 506 赞 · 51 转

New Gemini Audio models are available for developers to build more intelligent conversational experiences via the Gemini API and Google AI Studio. Today, we released new Gemini Live models in the

中文介绍 Google AI Studio 发布面向开发者的新 Gemini 音频模型:Gemini 3.8 Live 与 3.5 Transcribe,可通过 Gemini API 和 AI Studio 构建实时语音对话应用,主打更智能的会话体验。

Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking

@GoogleAIStudio · 203.4K 粉丝 · 35.1K 阅 · 585 赞 · 47 转

Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking are our most advanced live dialogue models yet. Major upgrades in intelligence and parallel reasoning make it more intuitive to collaborate and

中文介绍 Google AI Studio 发布 Gemini 3.8 Live 与 Gemini 3.8 Live Extended Thinking 两款实时对话模型,称其为迄今最先进的 live 模型,在智能水平与并行推理上有重大升级,让协作交互更自然。

AI Agents Are Moving Beyond Chat

@Didotxyz_ · 64.6K 粉丝 · 5.7K 阅 · 502 赞 · 325 转

The important shift in AI agents is not the disappearance of chat. It is that chat is no longer the whole product. An assistant explains a task. An agent system has to carry it forward: call a tool,

中文介绍 Didot 团队提出 AI 智能体的关键转变不是聊天消失,而是聊天不再等于整个产品:助手只负责解释任务,智能体系统则必须继续推进——调用工具、执行动作,把任务真正落地。

How Cooley is accelerating IPO work with ChatGPT

Cooley built GO Public with ChatGPT Work to bring intelligence to the IPO process, helping lawyers surface issues earlier and focus judgment where it matters most.

中文介绍 律所 Cooley 基于 ChatGPT Work 构建 GO Public 工具,将智能引入 IPO 流程,帮助律师更早发现问题,并把判断力集中在最关键环节。

Introducing Astra for Law

OpenAI for Law brings frontier intelligence for law, custom firm workflows, connected legal data sources, and legal-grade controls for confidential client work.

中文介绍 OpenAI 推出面向法律行业的 Astra for Law,提供前沿智能能力、律所定制工作流、连通的法律数据源,以及适用于保密客户工作的法律级管控。

Our framework for reporting model misalignment

OpenAI shares a framework for tracking, investigating, and disclosing model misalignment, alongside six reports of unexpected or concerning model behavior.

中文介绍 OpenAI 公布一套用于追踪、调查和披露模型失准的框架,并同时发布六份关于模型意外或令人担忧行为的报告。

Helping older adults use AI in everyday life

OpenAI and AARP are bringing free, hands-on ChatGPT workshops to 1,000 older adults across 10 U.S. cities to build practical AI skills safely.

中文介绍 OpenAI 与美国退休人员协会 AARP 合作,在美国 10 座城市为 1000 名老年人提供免费 ChatGPT 实操工作坊,帮助其安全掌握实用 AI 技能。

Reimagining advertising with AI

Explore new AI-powered advertising experiences from OpenAI, including Sponsored Agents, tools for marketers, and integrations with HubSpot and Shopify.

中文介绍 OpenAI 介绍 AI 驱动的广告新体验,包括赞助型智能体、面向营销人员的工具,以及与 HubSpot 和 Shopify 的集成。

Building the materials foundation for AI

The AI boom is becoming a materials challenge. As AI pushes computing into new territory, the materials behind that infrastructure are becoming just as crucial as the algorithms running on it. Semiconductors and data centers are approaching physical limits around performance, thermal management, ele

中文介绍 MIT 科技评论指出,AI 热潮正演变为材料挑战。随着算力向新领域推进,半导体和数据中心在性能等方面接近物理极限,支撑基础设施的材料与算法同等关键。

How to connect AI usage to business value

Learn how ChatGPT Work and Codex analytics help teams understand AI usage and spend, identify training needs, and connect adoption to business outcomes.

中文介绍 OpenAI 介绍 ChatGPT Work 与 Codex 的分析功能,帮助团队了解 AI 使用量与支出、识别培训需求,并将 AI 采用情况与业务成果挂钩。

How workers are unlocking new ways of working

New OpenAI Economic Research shows how workers use AI beyond traditional roles and which new activities become recurring parts of their work.

中文介绍 OpenAI 经济研究新报告显示,员工正以超出传统岗位的方式使用 AI,并揭示哪些新活动正成为其工作中的固定组成部分。

Can Skills Learned in Games Transfer to Real-World Work?

Good Start Labs trained an AI on a railroad game — and one version improved at financial research. The difference was the training design.

中文介绍 Good Start Labs 用一款铁路游戏训练 AI,其中一个版本在金融研究任务上表现提升,差异来自训练设计,探讨游戏技能能否迁移到现实工作。

Roundtables: Could AI really kill us all?

Listen to the session or watch below Employees at the world’s leading AI labs are saying there’s a real possibility that advanced AI could destroy humanity. Are they right? Or is this more scaremongering and hype? Watch a conversation unpacking AI extinction fears: where they come from, whether they

中文介绍 MIT 科技评论圆桌讨论:全球顶尖 AI 实验室员工称先进 AI 确有可能毁灭人类,与会者探讨这种 AI 灭绝风险说法是真实威胁还是危言耸听与炒作。

市场总览

美股技术面整体偏强:SPY 现价 762.6,位于 SMA20 764.67 下方但高于 SMA50 759.19 与 SMA200 715.48,多头排列成立;QQQ 单日 +1.76%,RSI14 分别报 41.8 与 44.1,动量温和。个股分化明显,AAPL 距 52 周高仅 -2.2%、5 日 +6.87%,META RSI14 达 72.2 处于超买,NVDA 则出现 MACD 死叉,TSLA 趋势中性。加密方面,恐慌贪婪指数 56 属贪婪档,总市值 2.63 万亿美元、24 小时 -1.98%,BTC 主导率 58.1%、ETH 11.3%;BTC 现价 76304.55 低于 SMA20 的 78008.95,但 SMA50 上穿 SMA200 构成金叉,ETH、SOL 均维持多头排列。中概股技术面整体承压,BABA、PDD、0700.HK 呈空头排列,JD RSI14 32.2 偏弱。商品与外汇分化:原油 RSI14 65.3 且多头排列,黄金中性;美元/人民币 RSI14 低至 19.9 处于超卖,美元指数与 10Y 美债收益率(4.95,RSI14 65.3)偏强。VIX 单日 -12.82% 至 15.44,处 52 周低位,风险指标读数平静。

今日关注

AAPL Apple (AAPL)
偏上行

现价 337,高于 SMA20 的 320.39 与 SMA50 的 319.27,多头排列成立;距 52 周高点仅 -2.2%,5 日累计 +6.87%,动量居美股前列。RSI14 报 62.1,位于中性偏强区间;MACD 4.6309 高于信号线 3.0924,红柱延续,趋势与动量读数方向一致。

META Meta (META)
中性

现价 682.31,较 SMA20 的 601.8 溢价明显,但趋势字段为 neutral,现价低于 SMA200 的 623.86 上方空间尚未确认。RSI14 高达 72.2 已进入超买区,MACD 21.8777 与信号线 11.8592 差距较大。超买与强势并存,方向读数偏中性。

NVDA Nvidia (NVDA)
中性

现价 219.34 微高于 SMA20 的 218.52,SMA50 为 213.58,多头排列仍在;但 MACD 0.4019 已低于信号线 2.0021,触发死叉,近 5 日 -1.94%。RSI14 报 46.7,略低于中轴。趋势偏多而动量转弱,呈多空交织的中性状态。

USDCNY=X 美元/人民币 (USDCNY=X)
偏下行

现价 6.7 贴近 52 周低点(距低仅 0%),并低于 SMA20 的 6.71、SMA50 的 6.74 与 SMA200 的 6.84,空头排列明确。RSI14 报 19.9 进入超卖区,MACD -0.0105 略高于信号 -0.0106 形成金叉。超卖与金叉或暗示跌势放缓,但排列仍偏下行。

BABA 阿里巴巴 (BABA)
偏下行

现价 108.54 低于 SMA20 的 115.06、SMA50 的 117.95 与 SMA200 的 133.78,标准空头排列。RSI14 报 33.7,逼近超卖阈值;MACD -3.2941 低于信号线 -2.487,绿柱扩张。距 52 周高点 -43.67%,1 日与 5 日均小幅收跌,技术面全面承压。

全部资产

^VIX

VIX 恐慌指数

$15.44 -12.82%
5 日
-13.45%
距 52w 高
-56.3%
RSI(14)
47.6
趋势
空头
SMA 20 / 50 / 200
15.72 / 16.18 / 18.10
MACD / 信号
0.240 / 0.065
空头排列

^TNX

10Y 美债收益率 (%)

$4.95 -1.18%
5 日
+0.06%
距 52w 高
-1.4%
RSI(14)
65.3
趋势
多头
SMA 20 / 50 / 200
4.81 / 4.71 / 4.39
MACD / 信号
0.084 / 0.067
接近 52 周高多头排列

DX-Y.NYB

美元指数 DXY

$100.23 -0.08%
5 日
+1.15%
距 52w 高
-1.5%
RSI(14)
60.6
趋势
多头
SMA 20 / 50 / 200
99.30 / 99.97 / 99.15
MACD / 信号
0.003 / -0.164
接近 52 周高多头排列

SPY

S&P 500 ETF

$762.60 +0.69%
5 日
+0.03%
距 52w 高
-2.2%
RSI(14)
41.8
趋势
多头
SMA 20 / 50 / 200
764.67 / 759.19 / 715.48
MACD / 信号
-0.736 / 1.205
接近 52 周高多头排列

QQQ

Nasdaq 100 ETF

$716.92 +1.76%
5 日
+0.09%
距 52w 高
-4.2%
RSI(14)
44.1
趋势
多头
SMA 20 / 50 / 200
712.67 / 709.88 / 661.59
MACD / 信号
-1.119 / 0.305
多头排列

AAPL

Apple

$337.00 +1.71%
5 日
+6.87%
距 52w 高
-2.2%
RSI(14)
62.1
趋势
多头
SMA 20 / 50 / 200
320.39 / 319.27 / 285.79
MACD / 信号
4.631 / 3.092
接近 52 周高多头排列

MSFT

Microsoft

$497.75 +0.13%
5 日
+1.24%
距 52w 高
-10.1%
RSI(14)
51.7
趋势
多头
SMA 20 / 50 / 200
496.21 / 460.20 / 431.54
MACD / 信号
8.122 / 11.598
多头排列

NVDA

Nvidia

$219.34 +3.38%
5 日
-1.94%
距 52w 高
-7.3%
RSI(14)
46.7
趋势
多头
SMA 20 / 50 / 200
218.52 / 213.58 / 197.95
MACD / 信号
0.402 / 2.002
MACD 死叉 (4 天前)多头排列

GOOGL

Alphabet

$347.33 +0.68%
5 日
+5.04%
距 52w 高
-15.0%
RSI(14)
50.0
趋势
多头
SMA 20 / 50 / 200
341.21 / 346.06 / 337.19
MACD / 信号
-1.595 / -2.568
MACD 金叉 (2 天前)多头排列

TSLA

Tesla

$366.20 +2.70%
5 日
-0.44%
距 52w 高
-26.6%
RSI(14)
51.1
趋势
中性
SMA 20 / 50 / 200
357.89 / 351.65 / 397.97
MACD / 信号
3.507 / 3.544
MACD 死叉 (今天)

META

Meta

$682.31 +1.80%
5 日
+4.38%
距 52w 高
-13.7%
RSI(14)
72.2
趋势
中性
SMA 20 / 50 / 200
601.80 / 605.11 / 623.86
MACD / 信号
21.878 / 11.859
RSI 超买
加密恐慌贪婪
56
贪婪
加密总市值
$2.63 T
-1.98% / 24h
BTC 主导率
58.1%
ETH 11.3%
24h 成交量
$77.8 B
活跃币 21,260

BTC-USD

Bitcoin

$76,304.55 +0.20%
5 日
-1.25%
距 52w 高
-39.5%
RSI(14)
50.7
趋势
多头
SMA 20 / 50 / 200
78,008.95 / 72,147.19 / 70,336.33
MACD / 信号
870.984 / 1,620.746
金叉(SMA50↑SMA200) (9 天前)多头排列

ETH-USD

Ethereum

$2,438.02 +0.91%
5 日
-3.48%
距 52w 高
-48.7%
RSI(14)
53.5
趋势
多头
SMA 20 / 50 / 200
2,463.82 / 2,227.94 / 2,067.34
MACD / 信号
51.893 / 77.145
多头排列

SOL-USD

Solana

$101.07 +2.46%
5 日
-0.70%
距 52w 高
-60.1%
RSI(14)
54.5
趋势
多头
SMA 20 / 50 / 200
101.82 / 89.94 / 83.41
MACD / 信号
2.193 / 3.433
多头排列

BABA

阿里巴巴 (BABA)

$108.54 -0.73%
5 日
-0.79%
距 52w 高
-43.7%
RSI(14)
33.7
趋势
空头
SMA 20 / 50 / 200
115.06 / 117.95 / 133.78
MACD / 信号
-3.294 / -2.487
空头排列

PDD

拼多多 (PDD)

$77.73 -0.45%
5 日
-1.12%
距 52w 高
-44.2%
RSI(14)
36.7
趋势
空头
SMA 20 / 50 / 200
83.18 / 85.40 / 96.77
MACD / 信号
-2.335 / -1.826
空头排列

JD

京东 (JD)

$26.64 -1.66%
5 日
-1.33%
距 52w 高
-27.7%
RSI(14)
32.2
趋势
中性
SMA 20 / 50 / 200
28.13 / 29.63 / 29.07
MACD / 信号
-0.798 / -0.720

0700.HK

腾讯控股 (0700.HK)

HK$426.00 -1.71%
5 日
+0.09%
距 52w 高
-37.6%
RSI(14)
40.1
趋势
空头
SMA 20 / 50 / 200
439.32 / 453.32 / 507.89
MACD / 信号
-6.571 / -6.183
空头排列

GC=F

黄金期货

$4,398.30 +0.25%
5 日
-0.20%
距 52w 高
-21.3%
RSI(14)
48.1
趋势
中性
SMA 20 / 50 / 200
4,499.34 / 4,319.54 / 4,552.59
MACD / 信号
-1.942 / 24.645

CL=F

WTI 原油期货

$100.91 -1.48%
5 日
-1.53%
距 52w 高
-15.5%
RSI(14)
65.3
趋势
多头
SMA 20 / 50 / 200
92.17 / 85.82 / 80.47
MACD / 信号
5.224 / 4.130
多头排列

USDCNY=X

美元 / 人民币

¥6.70 -0.15%
5 日
-0.15%
距 52w 高
-6.9%
RSI(14)
19.9
趋势
空头
SMA 20 / 50 / 200
6.71 / 6.74 / 6.84
MACD / 信号
-0.011 / -0.011
MACD 金叉 (4 天前)RSI 超卖接近 52 周低空头排列
风险提示

以上均为基于公开行情数据的技术指标读数复述,不构成任何投资建议或方向预测。过去走势不代表未来表现,指标可能滞后或失效,请结合自身情况独立判断,仅供技术指标解读参考。

Australia news live: RBA governor paves way for rate hike; Labor to ‘look at’ following Canada in pulling closer to EU

Follow the day’s news live Get our breaking news email, free app or daily news podcast Australia needs to slow down its artificial intelligence use and create better security laws, the companies creating the technology have told the federal government. In a submission to an ongoing inquiry, OpenAI s

中文摘要 澳大利亚储备银行行长释放加息信号;工党政府表示将研究是否效仿加拿大,与欧盟建立更紧密关系。同时,研发人工智能的企业向联邦政府提交意见,建议澳洲放缓AI应用步伐并完善安全法律。

Almost 50 dead in Nigeria after drinking alcohol suspected to contain methanol

Drinking locally made alcohol is common in Nigeria, especially in rural areas where it is more affordable and poorly regulated At least 48 people have died and about 100 others are receiving treatment in southern Nigeria after consuming a locally brewed alcoholic concoction suspected to contain meth

中文摘要 尼日利亚南部发生疑似甲醇中毒事件,至少48人死亡、约100人接受治疗。患者饮用了当地自酿的酒精饮品,该饮品疑似含甲醇。自酿酒在尼日利亚农村地区因价格低廉且监管薄弱而十分普遍。

Iran expels Swedish diplomat in retaliatory move

Tehran gives Swedish official 48 hours to leave as Iran says Stockholm is bowing down to outside pressure.

中文摘要 伊朗以对等反制方式驱逐一名瑞典外交官,并要求其在48小时内离境。伊朗方面称,瑞典是在向外部压力低头。

Iran war live: Trump weighs ‘big decision’ on Iran, tanker hit in Hormuz

Trump says he faces a 'big decision' on whether to 'annihilate' his Iranian foes, adding that 'anything could happen'.

中文摘要 特朗普表示正面临是否「消灭」伊朗对手的「重大决定」,并称「任何情况都可能发生」。与此同时,霍尔木兹海峡一艘油轮遇袭。

US court sentences ex-Syrian official who ran notorious ‘punishment wing’ to 60 years

Samir Ousman al-Sheikh, part of Bashar al-Assad’s regime, was in charge of Adra prison where prisoners were tortured This article contains descriptions of torture. A US federal court in Los Angeles sentenced a former Syrian government official who ran one of the country’s most infamous prisons to 60

中文摘要 美国洛杉矶联邦法院判处叙利亚阿萨德政权前官员萨米尔·乌斯曼·谢赫60年监禁。他曾主管以酷刑著称的阿德拉监狱「惩罚区」。报道提及内容包含酷刑描述。

Migrant crossings to Europe by sea drop sharply amid heightened risks: IOM

More than 2,200 migrants died or vanished at sea in 2026, despite declining crossings into Europe, says IOM.

中文摘要 国际移民组织(IOM)称,2026年经海路进入欧洲的移民人数大幅下降,但风险上升,当年已有超过2200名移民在海上死亡或失踪。

US approves $414M uranium investment in Niger, two years after troops left

Uranium project investment could mark a turning point in US-Niger and US-Canada relations.

中文摘要 美国批准向尼日尔铀矿项目投资4.14亿美元。此时距美军撤离该国已过去两年,该投资可能成为美国与尼日尔乃至美国与加拿大关系的转折点。

Yemen War Intensifies as Houthi Militia Advances on Marib

Clashes between Iranian-backed Houthi fighters and Saudi-backed government forces have been reported near oil-rich Marib Province.

中文摘要 也门战事升级,受伊朗支持的胡塞武装向石油资源丰富的马里卜省推进。据报道,胡塞武装与沙特支持的政府军在该省附近发生冲突。

YouTube star Ms Rachel to match Macklemore’s $1m Palestine aid donation

Educator talks of ‘moral obligation’ and urges ‘every wealthy white celebrity’ to make similar donation YouTube star Ms Rachel has said she is matching rapper Macklemore’s $1m donation to organizations supporting Palestinians, calling on other public figures to follow suit. Macklemore, who was fired

中文摘要 YouTube明星「Ms Rachel」表示将等额捐出100万美元,匹配说唱歌手Macklemore对支持巴勒斯坦组织的捐款。她称这是「道德义务」,并呼吁「每一位富有的白人名人」做出类似捐赠。

Russia election: Could other parties challenge United Russia?

Russians vote for the first time since 2022. Can opposition parties challenge Putin-backed United Russia in this vote?

中文摘要 俄罗斯举行自2022年以来的首次选举投票。报道关注反对党能否在此次投票中挑战普京支持的统一俄罗斯党。

Trinidad and Tobago ends state of emergency amid pushback over mass arrests

Lawyers criticised mass arrests under emergency powers that allowed authorities to detain people without charge.

中文摘要 特立尼达和多巴哥结束紧急状态。此前紧急权力允许当局未经指控拘留人员,引发大规模逮捕,律师对此提出批评。

US judge orders 30-day notice prior to any Kennedy Center physical changes

The ruling comes amid legal battle over Trump’s changes to the Kennedy Center, including his effort to rename it.

中文摘要 美国一名法官裁定,肯尼迪中心在进行任何实体改造前须提前30天通知。该裁决出台之际,围绕特朗普对肯尼迪中心的改动(包括更名尝试)正陷入法律争议。

US to allow Iran delegation to attend UN meetings in New York as war passes half-year mark

The BBC understands Iranian President Masoud Pezeshkian and Foreign Minister Abbas Araghchi are among the delegates allowed to enter the US.

中文摘要 在伊朗战争进入半年之际,美国将允许伊朗代表团赴纽约参加联合国会议。据BBC了解,获准入境美国的代表包括伊朗总统佩泽希齐扬和外长阿拉格齐。

Trump administration approves sale of F-35 jets to Saudi Arabia

The deal, which needs approval from Congress, comes as Riyadh seeks Washington's help in its war with Yemen's Houthis.

中文摘要 特朗普政府批准向沙特阿拉伯出售F-35战机,该交易尚需美国国会批准。此举出台之际,利雅得正寻求华盛顿在其与也门胡塞武装的战争中提供帮助。

Comedians Beyond Borders

A new generation of cross-cultural comics is taking regional humor global.

中文摘要 报道关注新一代跨文化喜剧演员,他们正将区域性幽默推向全球舞台。

El Niño Puts India on Track for Its Driest Monsoon Since 2009

India is on course for its driest monsoon in 17 years as a strengthening El Niño curbs rainfall, threatening crop yields and raising food-price risks.

中文摘要 受厄尔尼诺现象增强影响,印度今年季风雨量减少,或将迎来17年来最干燥的季风季。降雨不足威胁农作物产量,并推升食品价格上涨风险。

Vietnam Stocks See Early Foreign Inflows as FTSE Upgrade Nears

Global funds are dipping back into Vietnamese stocks ahead of their inclusion in FTSE Russell’s emerging markets universe next week.

中文摘要 在下周被纳入富时罗素新兴市场指数之前,全球基金开始重新流入越南股市。外资提前布局,押注指数升级带来的资金增量。

Gold Holds Gain as Lower Oil and Fed Hike Temper Inflation Fears

Gold held a gain at the end of a volatile week as the Federal Reserve’s first interest-rate hike since 2023 and lower oil prices helped ease concerns about inflation.

中文摘要 金价在震荡一周结束后保住涨幅。美联储自2023年以来首次加息,加之油价走低,缓解了市场对通胀的担忧。

Citi’s Korea Trading Head Says Buy Won Near 1,400 Per Dollar

The won’s recent weakness may be short lived as South Korea’s chip sector bolsters growth, supporting the currency’s recovery against the dollar by year-end, according to a Citigroup Inc. trader.

中文摘要 花旗集团一名交易主管表示,韩元近期疲软可能是短暂的。韩国芯片行业支撑经济增长,预计韩元将在年底前对美元回升,建议在1,400韩元兑1美元附近买入。

China's Gree Eyes Europe Winters for Growth

Gree Electric Appliances chairwoman Dong Mingzhu told Bloomberg TV's Avril Hong in a exclusive interview that Europe's cold winters represent the company's next growth opportunity, following record summer heat that nearly doubled air-conditioning revenue in the region. She also discussed Gree's Euro

中文摘要 格力电器董事长董明珠在接受彭博电视采访时表示,欧洲严寒冬季是公司下一个增长机会。此前欧洲夏季高温使该地区空调收入几乎翻倍。

RBA Must Judge If Rates High Enough to Cool Prices, Bullock Says

Australia’s central bank needs to decide later this month whether its benchmark interest rate is at a level sufficient to restrain price pressures and return inflation to target, Governor Michele Bullock said.

中文摘要 澳大利亚央行行长布洛克表示,央行需在本月晚些时候判断基准利率是否已高到足以抑制价格压力、使通胀回归目标水平。

Thames Water should be temporarily placed under public ownership, MPs warn

Move would be ‘cost neutral’ for the Treasury while protecting 16mn customers served by UK utility, cross-party parliamentary committee says

中文摘要 英国跨党派议会委员会警告,泰晤士水务应被暂时收归公有。该委员会称此举对英国财政部而言「成本中性」,同时可保护这家公用事业公司服务的1,600万用户。

UK losing £1.3mn a day on nationalised steelworks

MPs hit out at lack of plan for British Steel as ministers weigh new intervention in sector

中文摘要 英国议员批评政府对英国钢铁公司缺乏计划。报道称,国有化钢厂每天令英国损失130万英镑,部长们正考虑对该行业进行新的干预。

Early Bowie music available to stream for first time

Previously unheard material from the start of the artist's career has been released for the first time.

中文摘要 大卫·鲍伊职业生涯早期此前未公开的音乐素材首次开放流媒体播放。

UK Lawmakers Say Thames Creditors Have Been Joyriding Family Car

Thames Water’s senior creditors should be blocked from taking full control over the utility, a group of British lawmakers said Friday.

中文摘要 英国一群议员周五表示,应阻止泰晤士水务的高级债权人全面控制这家公用事业公司,并形容债权人此前如同「开着家里的车兜风」。

Why we bought our first home with a 100% mortgage - despite the risks

The share of UK mortgages with smaller deposits is the highest it's been since 2008. The BBC spoke to borrowers about how they manage the risks.

中文摘要 英国低首付房贷占比升至2008年以来最高。BBC采访了部分借款人以100%房贷购房的买家,了解他们如何应对相关风险。

The virtual worlds where robots are trained

Training systems that allow robots to negotiate the real world are getting more sophisticated.

中文摘要 用于训练机器人适应现实世界的虚拟世界系统正变得越来越复杂精细。

该源今日无内容。