官方新闻 · 2026-07
浏览 2026-07 发布的官方新闻内容，第 9 页，共 1197 条。
- 获联合国机构点赞！天立启鸣“AI+教育”方案入选AI for Good
- 国产开源新突破！5万小时真机数据，撕开行业最大痛点
- 竞赛编程Agent进入全球前十！南大、清华新模型CF rating超3500
- 刚刚，全球首个超高帧世界模型诞生！英伟达含量0，狂飙50帧
- DeepSeek之后，中国AI「自己出题」杀进Nature通讯！全球仅4家
- DeepSeek、智谱被曝造芯，国产大模型绕开英伟达！
- 智源悟界·RoboBrain Orca：多模态表征世界模型
- 小扎「芒果」生图只输GPT Image 2，没人教它改稿，它自己学会了
- 世界模型首次迎来“小时级”生成！蚂蚁灵波开源LingBot-World 2.0，支持AI原生多人交互
- 刚刚，Claude Fable 5续命5天！省钱攻略来了
- 今天，Claude Cowork大更新！合上电脑，它替你彻夜打工
- 蚂蚁灵波开源LingBot-Video，全球首个面向具身的视频基模来了！
- 蚂蚁灵波世界模型2.0正式发布！两大首创：小时级实时生成+Agent实时“陪玩”
- 蚂蚁灵波定义视频生成第二赛道，开源LingBot-Video，融合7万小时具身数据
- 融合7万小时具身数据，蚂蚁灵波开源具身视频基础模型LingBot-Video
- 从餐饮后厨到物理世界基础模型：量化派的物理AI，走的是卖能力的路
- WWW 2026 唯一最佳长文｜大模型该信「查到的」还是「记得的」？人大&腾讯提出知识需求驱动的新框架
- OpenAI深夜放出GPT-Live，ChatGPT终于像真人一样说话
- 同声传译一夜失业！GPT-Live瞬间翻译，老太太现场抬杠AI看傻全网
- Introducing a way to reflect on how you use Claude
- Inviting hard questions
- UST is bringing Claude to physical AI
- Why AI Infrastructure must evolve for Agent Experience — Akshat Bubna, Modal CTO
- Lovable reportedly in talks to double its valuation to $13.2B
- Conversations with AI: OpenAI's New Model
- Google’s deepfake detector system used to debunk McConnell hoax pic
- SpaceXAI releases Grok 4.5, which Elon describes as an ‘Opus-class model’
- This startup thinks robotics is about to have its ChatGPT moment
- Google Photos adds a new AI ‘Video Remix’ tool
- Why this CEO thinks video games make better training data than the internet
- Meta wants its AI glasses to seem less creepy. Its AI strategy says otherwise.
- ChatGPT’s upgraded voice mode is better at shutting up
- OpenAI releases new voice models for more natural live conversations
- Benchmarking Coding Agents on Databricks’ Multi-Million Line Codebase
- Prime Intellect raises $130M Series A to help enterprises build their own AI agents
- Flint: A visualization language for the AI era
- These AI startups are growing revenue at faster and faster rates
- NVIDIA Nemotron Achieves Benchmark-Leading Performance With LangChain Deep Agents Harness
- 机器人为什么要拟人？终于有人正确回答了
- 蒸馏效果起飞！DOPD破解「特权幻觉」，让在线策略蒸馏更有效
- Our approach to government and national security partnerships
- 50FPS、成本打掉70%，魔芯MoWorld把世界模型带进产业时代
- 具身智能“高考”难疯了！人类100分，最强模型12.8
- Your gaming data could be the secret to AGI, according to this Bezos-backed startup
- Separating signal from noise in coding evaluations
- Improving Best-of-N with Budget-Aware Execution for SWE Agents
- 横跨淘天与ATH事业群，硬核少年技术节下周京杭同步开幕，四项AIGX硬核成果齐发
- “人机共生，产需共融”——2026世界机器人大会新闻发布会在京召开
- Helping K–12 educators build practical AI skills
- 别让AI一上来就“进厂打螺丝”：智源悟界·Orca要先教模型理解世界如何变化
- 硬核少年技术节5.0京杭即将启幕，四项AIGX硬核成果发布，AI从工具走向伙伴
- 翁荔新博客提出「自进化先从Harness开始」，DeepSeek崔添翼转发附议
- Hot French startup ZML releases free product to speed inference across lots of AI chips
- 阿里斩获国际AI顶会最佳资源论文奖，提出Agent评测新范式
- 高德发布Phys AI Data：首个面向物理AI训练与应用的一站式空间数据基座
- 智能伙伴 共创未来！WAIC 2026即将举行并首发主题片
- 高德发布Phys AI Data：首个面向物理AI训练与应用的一站式空间数据基座
- 当AI开始重写自己：EverMind的自进化智能体框架Raven，正在定义数字生命的下一个形态
- DeepSeek被曝自研AI推理芯片
- DeepSeek秘密造芯！专攻推理，一年前已启动，招聘全程不公开
- 五位AI专家组成一支「虚拟科研团队」：Agentic AI开始独立设计临床试验，数周工作压缩至数分钟
- 论坛排片｜WAIC 2026最新论坛活动日程安排发布！
- 60000小时炼出新开源VLA！20多种机器人都能用
- 新闻发布会｜智能伙伴 共创未来！WAIC 2026即将举行并首发主题片
- 众智FlagOS获颁全球AI for Good优秀案例，推动AI算力普惠
- 突破具身智能泛化瓶颈！支持20多种机器人构型，蚂蚁灵波开源具身基座模型LingBot-VLA 2.0
- 20种机器人本体通吃！蚂蚁新一代VLA具身大脑刚刚开源了
- 中国科学院携手全球顶尖机构发布首个十亿参数级的手术视频基础模型SurgMotion！
- 提示词压缩竟成大模型新漏洞？港科大提出黑盒攻击框架COMA | ASE 2026
- 支持17家机器人厂商20多种构型，蚂蚁灵波LingBot-VLA 2.0正式开源
- 突破具身智能泛化瓶颈！支持20多种机器人构型，蚂蚁灵波开源具身基座模型LingBot-VLA 2.0
- 维纳智能登上Nature通讯：AI不只会回答问题，开始生成高精度行业数据
- RL训练一层就够了！单层RL超越全参数训练，跨任务跨模型跨算法全部验证
- [AINews] Lilian Weng summarizes 35 papers on Harness Engineering for RSI
- 三个理想火枪手创业，打破具身最快百台交付纪录
- 刚刚，Anthropic切开Claude大脑！AI自发长出类人「意识器官」
- 具身智能空间视觉死穴，终于被最新顶会彻底解决！
- 12项世界第一！国产基模一战封神，11亿参数逆袭70亿大魔王
- Fable 5手搓首个CUDA「超级内核」！2.5小时狂飙18.7倍
- OpenAI做不出的「中国式私教」！AI重构不可能三角，学而思超级大招
- 刚刚，全球首个加密AI产品完成保险签约！你敢用，它敢赔
- Claude Code惊人身世曝光！竟脱胎于安全对齐，Boris：才完成1%
- Open, convenient and predictable: Introducing Provisioned Throughput
- Introducing GPT-Live
- Meta rolls out Muse, a new AI image generator
- Claude's Mobile Co-Work Tools Are Here
- Meta’s new Muse Image model can pull other Instagram users into AI photos
- The Biggest AI Security Problem Isn't the Model. It's This. | Devvret Rishi
- Why the rise of open source AI isn’t hurting Anthropic … yet
- Microsoft joins AI cost-cutting trend by relying more on its own models
- Discord admits AI moderation bug wrongfully banned users over harmless images
- Anthropic is launching Claude Cowork on mobile and web
- Claude Cowork expands to mobile and web
- AI Innovators Adopt NVIDIA Vera — Why Max Single-Threaded CPU at Scale Matters
- Solos debuts an even lighter version of its camera-less smart glasses
- Introducing Robostral Navigate
- Savi’s app aims to protect consumers from realistic AI scams like kidnappers demanding ransom
- The foundational elements of AI architecture that IT leaders need to scale
- AI 只好看不好用？金山办公给出 “三通一平” 实战落地方法论
- 世界模型的DeepSeek时刻！魔芯Flash World Model降本70%，跑出50FPS实时交互