Guozhen AIGlobal AI field notes and model intelligence
中文

Realtime AI News

Track model releases, AI agents, open-source projects, infrastructure, policy, and product updates as they publish.

Read details: Ant Lingbo Open-Sources First Spatial-Native Embodied Vision Foundation Model
Ant Group / Embodied AI / Open Source

Ant Lingbo Open-Sources First Spatial-Native Embodied Vision Foundation Model

Ant Lingbo, Ant Group's robotics and embodied AI team, open-sourced what it calls the first spatial-native embodied vision foundation model on July 7, designed to help robots perceive and understand 3D spatial structures rather than just process 2D images. The model prioritizes spatial perception at the architectural level, potentially providing a new open-source foundation for embodied AI research and development.

Read details View source
蚂蚁灵波开源首个空间原生具身视觉基模，机器人三维感知能力有望跃升
Read details: Ant Group's Lingbo Tech Releases LingBot-Depth 2.0 and Open-Sources LingBot-Vision Foundation Model
蚂蚁集团 / 具身智能 / 空间感知

Ant Group's Lingbo Tech Releases LingBot-Depth 2.0 and Open-Sources LingBot-Vision Foundation Model

On July 7, Ant Group's embodied AI subsidiary Lingbo Technology officially released LingBot-Depth 2.0, a spatial perception model trained on 150 million samples that achieved 12 first-place results across 16 depth completion benchmarks. The company also open-sourced LingBot-Vision, a visual foundation model available in four versions (ViT-G/L/B/S).

Read details View source
Read details: WAIC 2026 Highlights Theoretical Breakthrough: Mathematical Bidirectional Empowerment Opens New AI Paradigm
WAIC / AI Theory / China

WAIC 2026 Highlights Theoretical Breakthrough: Mathematical Bidirectional Empowerment Opens New AI Paradigm

At WAIC 2026, theoretical breakthroughs took center stage with the concept of ‘mathematical bidirectional empowerment’ proposed to drive AI paradigm innovation. The approach aims to leverage mutual reinforcement between mathematics, physics, and AI to explore foundations beyond current deep learning paradigms.

Read details View source
Read details: First AI Execution Hub for Home Furnishing: SVJ AI Workbench Officially Launches
三维家 / AI执行中枢 / 家居行业

First AI Execution Hub for Home Furnishing: SVJ AI Workbench Officially Launches

Sanweijia (SVJ) has launched what it calls the first AI execution hub for the home furnishing industry, the SVJ AI Workbench. The platform represents a shift from AI as a design assistant to AI as an autonomous execution core for the full workflow from design to production.

Read details View source
Read details: Alipay Launches AI Open Platform: One Integration, Multi-Device Distribution Across Phones, Cars, and Smart Glasses
支付宝 / AI开放平台 / Agent

Alipay Launches AI Open Platform: One Integration, Multi-Device Distribution Across Phones, Cars, and Smart Glasses

Alipay officially launched its AI Open Platform on July 7, opening AI integration capabilities to merchants, institutions, and service providers. Businesses can connect to Alipay's AI assistant 'Abao' and distribute services across smartphones, car systems, AI glasses, and IoT devices with a single integration, marking a shift from search-based to conversational service delivery.

Read details View source
Read details: Tencent Launches AI Agent Red Teaming Framework, First to Audit MCP Supply Chain Security
Tencent / AI Security / MCP

Tencent Launches AI Agent Red Teaming Framework, First to Audit MCP Supply Chain Security

Tencent has released an AI agent red teaming framework that, for the first time, extends security auditing to the MCP (Model Context Protocol) supply chain. The framework is designed to detect and prevent supply chain attack vectors introduced through MCP, providing a systematic assessment tool for secure agent ecosystem deployment.

Read details View source
腾讯首发AI智能体红队测试框架，首次审计MCP供应链安全
Read details: Volvo EX60 Grants Gemini Camera Access, Bringing AI Vision to Vehicles
沃尔沃 / Gemini / 车载AI

Volvo EX60 Grants Gemini Camera Access, Bringing AI Vision to Vehicles

Volvo has announced that its EX60 electric SUV will grant Google's Gemini AI assistant access to its onboard cameras, giving the AI real-time visual perception of the vehicle's environment. This marks a significant step toward context-aware automotive AI assistants that can see and understand the world around the car.

Read details View source
Read details: Anthropic Discovers Claude's Hidden 'J-Space' — A Brain-Like Region Whose Removal Cripples Reasoning
Anthropic / Claude / Interpretability

Anthropic Discovers Claude's Hidden 'J-Space' — A Brain-Like Region Whose Removal Cripples Reasoning

Anthropic has identified a hidden representational region inside Claude's neural network called 'J-Space,' which exhibits structured characteristics resembling human cognitive maps. Researchers found that removing or disrupting J-Space caused Claude's reasoning capabilities to degrade sharply, offering an unprecedented window into how large language models organize internal knowledge and perform logical operations.

Read details View source
Anthropic发现Claude内部「J-Space」类脑表征空间：删除后推理能力急剧下降
Read details: Virtual telcos spark AI price war in Singapore: $8/month bundles pack multiple premium AI models
Circles.Life / giga! / AI

Virtual telcos spark AI price war in Singapore: $8/month bundles pack multiple premium AI models

Singapore's virtual telcos Circles.Life and giga! are bundling access to paid AI models with mobile plans starting as low as $8 per month, targeting youths aged 15-29. The plans aggregate ChatGPT, Gemini, Claude and others through proprietary apps, marking a new front in telecom differentiation.

Read details View source
Read details: Hugging Face and SkyPilot launch zero-egress cloud storage for AI workloads across any cloud
Hugging Face / SkyPilot / 云存储

Hugging Face and SkyPilot launch zero-egress cloud storage for AI workloads across any cloud

Hugging Face has partnered with open-source multi-cloud orchestration platform SkyPilot to offer zero-egress storage, enabling users to run AI workloads on any cloud while storing data directly on Hugging Face. The integration addresses the high cost of data transfer in multi-cloud environments, giving AI teams greater architectural flexibility.

Read details View source
Hugging Face 联合 SkyPilot 推出零出站流量云存储方案，AI 工作负载可跨任意云运行
Read details: MUFG partners with OpenAI to become an AI-native financial organization
MUFG / OpenAI / Financial Services

MUFG partners with OpenAI to become an AI-native financial organization

Mitsubishi UFJ Financial Group (MUFG), Japan's largest banking group, is deploying ChatGPT Enterprise to build an AI-native organization and deliver AI-powered financial services at scale. The move signals a strategic shift among traditional financial giants from AI experimentation to institutional deployment.

Read details View source
三菱日联金融集团宣布与OpenAI合作，打造AI原生金融机构
Read details: Aureka Launches OpenDDE: An Open-Source Drug Discovery Engine to Accelerate AI-Driven Therapeutics
Drug Discovery / Open Source / AI

Aureka Launches OpenDDE: An Open-Source Drug Discovery Engine to Accelerate AI-Driven Therapeutics

Aureka has released OpenDDE, an open-source drug discovery engine designed to accelerate AI-driven therapeutic development. The project aims to lower the technical barrier for AI-powered pharmaceutical research by making core tools openly accessible to the global research community.

Read details View source
Read details: Sysdig Uncovers First Fully AI-Agent-Driven Ransomware Attack: JADEPUFFER
Sysdig / Ransomware / AI Agent

Sysdig Uncovers First Fully AI-Agent-Driven Ransomware Attack: JADEPUFFER

Sysdig has documented the first known ransomware campaign driven end-to-end by an AI agent's own decision-making. The JadePuffer operation executed over 600 distinct payloads autonomously, with the AI redeploying a corrected payload just 31 seconds after hitting an error.

Read details View source
Read details: Bitcoin miner TeraWulf shares surge on $19B Anthropic AI infrastructure lease deal
Anthropic / TeraWulf / AI Infrastructure

Bitcoin miner TeraWulf shares surge on $19B Anthropic AI infrastructure lease deal

Bitcoin mining company TeraWulf saw its share price rise after signing a $19 billion AI infrastructure lease with Anthropic and completing a joint venture sale. The deal underscores the accelerating pivot of crypto mining infrastructure toward AI compute services.

Read details View source
Read details: Sam Altman Pushes OpenAI Equity Plan: Each American Family Could Get a $300 Stake
OpenAI / Sam Altman / AI Economy

Sam Altman Pushes OpenAI Equity Plan: Each American Family Could Get a $300 Stake

OpenAI CEO Sam Altman is advancing a plan to distribute AI-generated wealth to Americans, with each family potentially receiving roughly $300 in OpenAI equity, according to a Financial Times report. MIT Technology Review analyzes the proposal as more significant as a political narrative than as a concrete policy plan.

Read details View source
Sam Altman推进OpenAI全民股权计划：每个美国家庭或获300美元权益
Read details: Cloudflare launches stablecoin payment gateway for AI agents — Monetization Gateway
Cloudflare / Stablecoin / AI Agent

Cloudflare launches stablecoin payment gateway for AI agents — Monetization Gateway

Cloudflare has opened a waitlist for its Monetization Gateway, which lets customers charge for any resource on Cloudflare's network using stablecoin settlements. The infrastructure enables AI agents to autonomously pay for services in machine-to-machine transactions.

Read details View source
Cloudflare推出AI Agent稳定币支付方案——Monetization Gateway
Read details: Alibaba Bans Employees From Using Anthropic's Claude Code Over Security Concerns
Alibaba / Anthropic / Claude Code

Alibaba Bans Employees From Using Anthropic's Claude Code Over Security Concerns

Alibaba has prohibited its employees from using Anthropic's Claude Code AI coding assistant due to security concerns. The ban reflects growing caution among Chinese tech giants about exposing proprietary codebases to third-party AI tools based in the United States.

Read details View source
Read details: NVIDIA Dominates ICML 2026: 74 Papers Accepted, 145 Cite Open Nemotron Models
NVIDIA / ICML / Open Source

NVIDIA Dominates ICML 2026: 74 Papers Accepted, 145 Cite Open Nemotron Models

NVIDIA had 74 papers accepted at ICML 2026, with approximately 2,000 accepted papers citing NVIDIA GPUs and 145 citing NVIDIA Nemotron open models as foundational research tools. The findings confirm that open frontier models and open AI infrastructure have become essential to modern AI science.

Read details View source
NVIDIA在ICML 2026展示开放模型主导力：74篇论文入选，145篇引用Nemotron
Read details: SurrealDB unveils high-speed AI agent context layer, merging multi-model database with agent infrastructure
SurrealDB / AI Agent / Database

SurrealDB unveils high-speed AI agent context layer, merging multi-model database with agent infrastructure

Multi-model database startup SurrealDB has showcased its high-speed context layer for AI agents, built on a custom storage engine that unifies relational, graph, vector, and document queries in a single system. Verizon, Tencent, and Samsung Ads are already using the technology in production.

Read details View source
SurrealDB推出面向AI Agent的高速上下文层，数据库与AI代理深度融合
Read details: Inside the Secret AI War Between Silicon Valley and China: Washington Post Investigation
Washington Post / US-China / AI Policy

Inside the Secret AI War Between Silicon Valley and China: Washington Post Investigation

The Washington Post published a major investigative report revealing the hidden competition between Silicon Valley and China over artificial intelligence. The deep-dive story exposes the strategic maneuvering, talent wars, and technology rivalry between the two tech ecosystems.

Read details View source
Read details: First Fully Autonomous Ransomware Attack Executed Entirely by an AI Agent
AI Agent / Security / Ransomware

First Fully Autonomous Ransomware Attack Executed Entirely by an AI Agent

Security researchers at Sysdig have identified the first fully autonomous ransomware attack conducted by an AI agent with no human involvement. The JadePuffer ransomware operation used an autonomous LLM agent to perform the entire attack chain — from reconnaissance and vulnerability exploitation to credential theft, lateral movement, and file encryption.

Read details View source
AI Agent首次自主实施完整勒索攻击：从漏洞扫描到加密文件全程无人干预
Read details: Station F Ramps Up F/ai Accelerator, Doubling Down on Europe's AI Startup Ecosystem
Station F / AI Accelerator / Europe

Station F Ramps Up F/ai Accelerator, Doubling Down on Europe's AI Startup Ecosystem

Station F, the landmark Parisian startup campus founded by billionaire Xavier Niel, is preparing a new edition of its F/ai accelerator program to strengthen its role as a launchpad for Europe's most promising AI startups. The move signals a strategic push to capture the growing wave of AI entrepreneurship across the continent.

Read details View source
巴黎Station F升级F/ai加速器，剑指欧洲AI创业第一站
Read details: Beijing Unveils 'AI4S 18 Measures' to Deploy AI Scientists and Build 24-Hour Unmanned Labs
北京 / AI4S / 政策

Beijing Unveils 'AI4S 18 Measures' to Deploy AI Scientists and Build 24-Hour Unmanned Labs

Beijing has officially released 18 policy measures for AI for Science (AI4S), including initiatives to deploy AI scientists in research roles and construct 24-hour unmanned laboratories. The policy package aims to deeply integrate AI with fundamental scientific research and accelerate Beijing's emergence as a global AI4S innovation hub.

Read details View source
Read details: Tencent's Hunyuan 3 Release Drives Stock Price Up, Intensifying Pressure on Zhipu and MiniMax
Tencent / Hunyuan 3 / Stock Market

Tencent's Hunyuan 3 Release Drives Stock Price Up, Intensifying Pressure on Zhipu and MiniMax

Tencent's official release of Hunyuan 3 has driven its stock price higher while putting greater competitive pressure on rivals including Zhipu and MiniMax. The fully open-source model with integrated Agent capabilities is reshaping China's large language model landscape.

Read details View source
腾讯混元3正式发布推动股价上涨，智谱与MiniMax面临竞争压力
Read details: Chinese AI Firm Zhiping Unveils Brain-Inspired NeuroVLA Model at UN, Offering a Low-Power Alternative to Compute-Intensive Robotics AI
智平方 / 具身智能 / 类脑大模型

Chinese AI Firm Zhiping Unveils Brain-Inspired NeuroVLA Model at UN, Offering a Low-Power Alternative to Compute-Intensive Robotics AI

Zhiping Technology, a Chinese embodied AI company from the Greater Bay Area, showcased its proprietary brain-inspired large model NeuroVLA and the open-source AlphaBrain Platform at the UN Open Source Week in New York. The model draws on a cortex-cerebellum-spinal cord collaborative mechanism, cutting robot motion jitter by over 75% while consuming as little as 0.4 watts at the spinal layer.

Read details View source
Read details: First Vertical LLM for Foreign Language Discipline 'Luyan' Officially Released
麓言 / 垂直大模型 / 外语教学

First Vertical LLM for Foreign Language Discipline 'Luyan' Officially Released

The first vertical large language model built specifically for foreign language education and research, named 'Luyan,' has been officially released in China. The model targets the unique needs of foreign language teaching, translation, and linguistic research with customized optimization.

Read details View source
Read details: ByteDance, Unitree, BAT Product Leaders Headline 2026 Singularity Intelligent Product Conference in Beijing
奇点智能产品大会 / Agent / AI产品

ByteDance, Unitree, BAT Product Leaders Headline 2026 Singularity Intelligent Product Conference in Beijing

The 2026 Singularity Intelligent Product Conference, co-organized by CSDN and the Singularity Intelligent Research Institute, announced its full agenda on July 6 with over 40 product leaders from ByteDance, Unitree, Baidu, Alibaba, and Tencent. The two-day event in Beijing will cover Agent AI, enterprise AI deployment, AI Coding, embodied intelligence, and AI-native organizations.

Read details View source
Read details: China's CAC Completes Phase 1 of 'Qinglang·AI Application Chaos' Crackdown, Targets Fake News and Minors' Protection in Phase 2
中央网信办 / AI监管 / 政策

China's CAC Completes Phase 1 of 'Qinglang·AI Application Chaos' Crackdown, Targets Fake News and Minors' Protection in Phase 2

China's Cyberspace Administration (CAC) announced it has completed the first phase of its special campaign to rectify chaos in AI applications. Phase 2 will focus on AI-generated fake information, violent content, impersonation, infringement on minors' rights, and AI-powered astroturfing operations.

Read details View source
Read details: Tencent Officially Releases Hunyuan Hy3 with Open-Source Access, Nearly 50 Internal Units Queue for Integration
Tencent / Hunyuan Hy3 / Model Release

Tencent Officially Releases Hunyuan Hy3 with Open-Source Access, Nearly 50 Internal Units Queue for Integration

Tencent officially released Hunyuan Hy3 on July 6, open-sourcing the model and launching its Agent capabilities on Yuanbao for free public access. The input price is as low as 1 yuan per million tokens, and nearly 50 internal business units are already queuing to integrate the model.

Read details View source
腾讯混元 Hy3 正式发布并开源，近 50 个内部业务排队接入
Read details: Zhipu AI Stock Plunges 16% as Tencent's Hy3 Model Sparks Competition Fears
智谱 / Zhipu AI / 腾讯

Zhipu AI Stock Plunges 16% as Tencent's Hy3 Model Sparks Competition Fears

Shares of Hong Kong-listed AI company Zhipu AI plunged as much as 16.2% in afternoon trading after Tencent released its Hy3 large language model, triggering market concerns over intensifying competition in China's AI sector. Investors worry that Tencent's entry into the foundation model race could squeeze the market share of third-party AI companies.

Read details View source
腾讯发布Hy3模型引发竞争担忧，智谱港股午后跳水一度跌超16%
Read details: ByteDance's Seedance Is Quietly Taking Over Hollywood: $9/Minute AI Video Generation Wins Filmmakers
字节跳动 / Seedance / 视频生成

ByteDance's Seedance Is Quietly Taking Over Hollywood: $9/Minute AI Video Generation Wins Filmmakers

ByteDance's AI video generation tool Seedance is rapidly gaining traction in Hollywood, with costs as low as $9 per minute for video-plus-audio generation — far below Google Veo's $24. Producers are using it openly and secretly, with some studios receiving $2 million offers for unrestricted access.

Read details View source

Daily Archive

Daily Archive

Full Archive
2026-07-16Daily AI Archive | 2026-07-16This issue collects 23 AI updates from 2026-07-16, led by 美国AI新秀首发模型，参考中国技术|DeepSeek V3|Kimi K2.5|OpenAI|米拉·穆拉蒂|金融时报 手机新浪网 新浪财经, Applied Computing wants to give oil and gas operators an AI model for the entire plant, NVIDIA Introduces New Jetson Thor Computers to Advance Mainstream Robotics and Edge AI, Agentic orchestration: Enterprise AI organizations have a deployment problem, not a platform problem — and most are calling chatbots agents.2026-07-15Daily AI Archive | 2026-07-15This issue collects 31 AI updates from 2026-07-15, led by Meta's Adam Mosseri predicts AI token budgets may soon be capped per engineer, Microsoft Deploys In House MAI Models in Office Apps, Signaling Shift from Frontier AI to Deployment Economics, DeepMind CEO Hassabis calls for independent AI standards body modeled after CERN and IPCC, Google localises Gemini AI for regulated sectors in India, expands enterprise push with offline deployment.2026-07-14Daily AI Archive | 2026-07-14This issue collects 42 AI updates from 2026-07-14, led by Intel Invests €5 Billion to Expand Ireland Fab, Betting on AI Infrastructure Demand, Apple and OpenAI Trade Lawsuits Over AI Device Tech Theft, as China's StepX Neo Debuts AI Phone, Step Launches World's First Agent Native OS, Rebuilt from Scratch for AI Agents, iFlytek Open Sources Enterprise Agent Platform Under Apache 2.0, Reaches 8.6K GitHub Stars.2026-07-13Daily AI Archive | 2026-07-13This issue collects 23 AI updates from 2026-07-13, led by LTM Partners with Anthropic, Betting on Claude to Drive Enterprise AI Transformation, Apple Files Lawsuit Against OpenAI, Musk Altman Feud Erupts into Public War of Words, Birdsview Raises €2.5 Million Seed to Scale Avys AI Retention Agent for Retailers, SpaceXAI Targets Grok 4.5 at Developers and Financial Firms, Taking a Differentiated Commercial Path.2026-07-12Daily AI Brief — July 12, 2026: Mathematical Breakthroughs, Agent Tooling Explosion, and Tech Titan ManeuveringJuly 12 saw major breakthroughs in AI reasoning as OpenAI's GPT-5.6 Sol Ultra solved a 50-year-old graph theory conjecture. Apple sued OpenAI, sparking a public feud between Musk and Altman, while Microsoft accelerated in-house AI development to reduce dependence on external models. On-device AI and video generation also advanced, with StepFun launching edge models and ByteDance's Seedance 2.5 achieving native 30-second video generation.2026-07-11Daily AI Archive | 2026-07-11This issue collects 21 AI updates from 2026-07-11, led by SK Hynix Raises $26.5B in Largest Foreign IPO in US History, Micron Makes Strategic Investment in Anthropic, OpenAI's GPT 5.6 Sol autonomously post trained a smaller model from a 'fairly underspecified prompt', Apple Sues OpenAI Over Alleged Trade Secret Theft Directed by Senior Leadership.2026-07-10Daily AI Archive | 2026-07-10This issue collects 21 AI updates from 2026-07-10, led by OpenAI 最强 AI 模型：GPT 5.6 系列正式上线，纳德拉称微软 Copilot 同步接入 新浪网, Meta's Muse Image AI Generator Uses Public Instagram Photos by Default, Sparking Privacy Backlash, Meta's Next Gen AI Chips to Begin Production in September, Designed with Broadcom, Manufactured by TSMC, Paris AI Voice Startup Gradium Raises $100M Seed Round Backed by Nvidia.2026-07-09Daily AI Archive | 2026-07-09This issue collects 39 AI updates from 2026-07-09, led by NVIDIA Releases 'Data for Agents' Open Dataset on Hugging Face, Pressure Mounts for AI Regulation Focused on Children's Safety, OpenAI to Release GPT 5.6 Sol, Terra, and Luna on Thursday Amid Contradictory US Regulatory Signals, Google Photos Launches AI Powered Video Remix Tool Powered by Gemini Omni.2026-07-08Daily AI Archive | 2026-07-08This issue collects 19 AI updates from 2026-07-08, led by Anthropic Expands Claude Cowork to Mobile and Web, Agent Use Data Reveals Business Operations as Top Use Case, Discord Admits AI Moderation Bug Wrongfully Banned Users Over Harmless Images, Microsoft Joins AI Cost Cutting Trend by Relying More on Its Own Models, Hugging Face Launches One Click Deployment to Amazon SageMaker Studio.2026-07-07Daily AI Archive | 2026-07-07This issue collects 36 AI updates from 2026-07-07, led by NVIDIA Dominates ICML 2026: 74 Papers Accepted, 145 Cite Open Nemotron Models, Alibaba Bans Employees From Using Anthropic's Claude Code Over Security Concerns, Google quietly expands AI training to include more user search data, Cloudflare launches stablecoin payment gateway for AI agents — Monetization Gateway.