AI daily

Daily archive

Browse published AI daily reports in reverse chronological order.

2026-08-09

2026-08-09 AI Daily

TBSM: A New Training Method for One-Step Generative Models Without Complex Distillation / Five flaws in the 'augmented work' myth: What a major Québec survey found / MCP's Biggest Update Yet: Back to the HTTP Era / 5 flaws in the 'augmented work' myth: What a major Québec survey found / U.S. DOE Launches Genesis Open Models Initiative to Develop Scientific AI Models

Selected

14

Sources

8

2026-08-08

2026-08-08 AI Daily

Anthropic Makes Auto Mode Default for Claude Code, Citing Better Safety Than Human Approval / DeepSeek Raises Prices 30-Fold, Yet Remains Cheapest; Third-Party Platforms Offer Even Lower Rates / Anthropic Details Claude's Security Isolation Architecture: Constraining Agent Behavior via Infrastructure / Fudan Team Develops Intraoperative 3D Pathology Platform with Label-Free Imaging in 30 Minutes / Floatboat Harness Benchmark: 1/57 Cost Beats Claude Opus 4.8 Across All Five Tests

Selected

10

Sources

5

2026-08-07

2026-08-07 AI Daily

Yuanli Lingji Releases Embodied Foundation Model DM0.5 to Accelerate Deployment / Improving Fable 5 Safeguards / Handroid: A Reconfigurable Robot That Is Both a Dexterous Hand and a Humanoid / Anthropic Details Claude's Security Isolation Architecture to Constrain Agent Behavior / Google AI Earthquake: Loses Four Core Scientists in One Day, the Restructuring Path of a $4.6 Trillion Giant

Selected

25

Sources

12

2026-08-06

2026-08-06 AI Daily

Google's AI Model WeatherNext Published in Nature, Boosts Cyclone Forecast Accuracy / WeatherNext: AI Model Achieves Breakthrough in Cyclone Forecasting / Sand.ai Open-Sources MAGI-2, the First 100B+ MoE Video Model, Slashing Costs / Alibaba Cloud Launches Wan3.0 Video Generation Model in Public Beta / Exclusive: Mirendil signs $100M+ Google Cloud deal to scale self-improving AI

Selected

46

Sources

24

2026-08-04

2026-08-04 AI Daily

Shieldstral Introduces 3B Multimodal Safety Classifier / Deploy local agents everywhere with LFM2.5-2.6B / NVIDIA Releases Alpamayo 2 Super: An Open Reasoning Vision Model to Unify Autonomous Vehicle Development / Tino Cuellar Joins Anthropic as Chief Global Affairs Officer / NVIDIA Alpamayo 2 Super, the Frontier Open Model for Robotaxis and Autonomous Vehicles, Now Available for Commercial Use

Selected

19

Sources

11

2026-08-03

2026-08-03 AI Daily

Alibaba Releases Qwen3.8: 2.4T Parameters, Enhanced Coding Abilities / AI Migrates Legacy COBOL Programs to Java, Bugs Included / EU Begins Enforcing AI Act Transparency Rules / EU AI labeling and transparency rules now in effect / NVIDIA Vera Storage Benchmarks: Faster Encryption, Compression, Integrity Checking, and Recovery for AI-Native Storage

Selected

20

Sources

12

2026-08-02

2026-08-02 AI Daily

Tsinghua Team Open-Sources VeriLoop Coder-E1: Evidence-Governed Spiral for Verifiable Recursive Self-Improvement / Generative AI Floods and Dilutes the Book Market / OpenAI's New Model Solves 10 Math Problems in a Row, Turning Math into a Click Game for $2000 / OpenAI's Next-Gen Astra Model Solves Ten Open Math Problems for Just $2,000 / MCP Gets Its Biggest Update: Back to the HTTP Era

Selected

8

Sources

6

2026-08-01

2026-08-01 AI Daily

OpenAI announces next major model Astra, releases ten previously unsolved math problems / EU to Mandate Labels on Authentic-Looking AI Content Starting August 2 / Generative Models Can Now Be Trained End-to-End? The Key Is a For Loop / Google, Microsoft, Amazon Earnings: AI Spending Strains Cash Flow, Custom Chips Shine / DeepSeek Launches V4-Flash Public Beta API, Claims Agent Capabilities Surpass V4-Pro-Preview

Selected

6

Sources

6

2026-07-30

2026-07-30 AI Daily

Legendary AlphaFold Team Disbanded: Nobel Laureate Joins Anthropic, Resources Shift to Gemini / Gemini Robotics ER 2: Powering Robotics with Video Understanding, Task Orchestration, and Multi-Robot Collaboration / Lilian Weng Returns to OpenAI One Day After Leaving Thinking Machines Lab / OpenAI Reveals Secret: Two Settings Triple GPT-5.6 Sol's ARC-AGI-3 Score / Anthropic Discloses Claude Model's Unauthorized Access to Real Systems During Evaluations

Selected

17

Sources

20

2026-07-27

2026-07-27 AI Daily

Kimi K3 Open-Source Weights Released / Terence Tao Warns of a New Century-Long Crisis in Mathematics at Fields Medal Ceremony / Anthropic CEO's Position on Open-Weights Models / Nvidia and Microsoft launch open AI security alliance, without OpenAI, Google, or Anthropic / Reimagining Independence: How Meta's AI Models Are Helping the University of Pittsburgh Transform Assistive Robotics

Selected

22

Sources

19

2026-07-26

2026-07-26 AI Daily

Opus 5 Surpasses Fable 5: Anthropic's Pricing Strategy Disrupted? / Tencent Open-Sources Three Embodied Foundation Models; Chief Scientist Zhang Zhengyou Explains How 'Three-Layer Brain' Solves Robot Latency / NeoteAI and Fudan University Release Tactile Data Technical Reports to Enhance Embodied AI's Sense of Touch / Anthropic's Opus 5 blows past Fable 5 and GPT-5.6 Sol on the benchmark designed to measure real intelligence / Valkor, Zhejiang University, and UCL Open Source Loom: A Structured Engineering State Layer for AI Coding Agents

Selected

6

Sources

16

2026-07-25

2026-07-25 AI Daily

Anthropic Launches Opus 5: Matches Fable 5 Performance at Half the Price / Anthropic Cuts 80% of Claude Code System Prompts as New Models Need Less Guidance / XYZ AI Lab Releases Two Deep Search Agents, Achieving Multiple SOTAs and Exploring AI4AI Paradigm / Running a 28.9M parameter LLM on an $8 microcontroller / Towards Long-Horizon Agents: Gaoling Institute Releases 149-Page Comprehensive Survey

Selected

9

Sources

16

2026-07-23

2026-07-23 AI Daily

ChatGPT Launches Health Feature / Google Releases Gemini 3.6 Flash and Two Other Models / Grok Build Launches Workflows Feature / ChatGPT Voice arrives in desktop app / Black Forest Labs Releases and Open-Sources Flux 3 Video Generation Model

Selected

16

Sources

17

2026-07-22

2026-07-22 AI Daily

Anthropic's $1.5B Piracy Settlement with Book Authors Is a Record Loss That Hands AI Labs Their Biggest Legal Win / OpenAI Claims Responsibility for Hugging Face Hack After Models Escape Sandbox / OpenAI Officially Responds to GPT-5.6 Sol Breaching Production Systems: The Era of Autonomous AI Attacks Has Arrived / Trump Administration Announces More Than $5B for the Genesis Mission, a National Mission on AI for Science / OpenAI Introduces Presence Platform

Selected

17

Sources

19

2026-07-21

2026-07-21 AI Daily

Google Introduces New Gemini Models: 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber / Alphabet Embeds Gemini Directly into Silicon / Judge Grants Final Approval of Anthropic's $1.5B Copyright Settlement With Authors and Publishers / NVIDIA Vera Rubin Driving Performance Per Watt, Lowest Token Cost for Partners Worldwide / Inside NVIDIA Rubin GPU Architecture: Powering the Era of Agentic AI

Selected

8

Sources

17

2026-07-20

2026-07-20 AI Daily

Google's 'Frozen v2' Chip Reportedly Bakes Gemini Architecture Directly into Silicon for Efficiency Gains / Bloomy (YC S26) launches AI-powered mastery learning for K-12 / Integrate NVIDIA Omniverse RTX Sensor Simulation Into Existing Apps / Ollama Secures $65M Series B Funding to Grow its Open-source AI Platform / Kapture CX Raises $10M Pre-Series B Led by Bajaj Finserv Ventures to Scale Agentic Enterprise Stack

Selected

6

Sources

6

2026-07-19

2026-07-19 AI Daily

Alibaba Launches Token Plan Personal Edition and Qwen3.8-Max Preview

Selected

1

Sources

2

2026-07-18

2026-07-18 AI Daily

Anthropic Makes Claude Fable 5 Permanent in Subscription Plans / China's new World Artificial Intelligence Cooperation Organization is President Xi's clearest play yet for a parallel AI order

Selected

2

Sources

8

2026-07-17

2026-07-17 AI Daily

Kimi K3 2.8T-A50B: Largest Open Model Ever Released, Opus 4.8-Class at Sonnet 5 Pricing / Thinking Machines Lab Launches Open-Weight AI Model Inkling, Betting on Customizable AI / Musk Open-Sources Grok Build Overnight, But Code Still Contains Traces of Uploading Users' Entire Repositories / Kimi K3 Sets Record with 2.8 Trillion Parameters in Open-Source Model / Moonshot AI Releases Kimi K3: World's First 3 Trillion Parameter Open Model

Selected

10

Sources

25

2026-07-16

2026-07-16 AI Daily

xAI Open-Sources 'Grok-Build' on GitHub After Massive Data Breach / Ex-OpenAI CTO Murati's Thinking Machines Drops Inkling, a 975B Parameter Model That Leads US Labs but Trails China / Kimi's Open Model K3 Nears GPT-5.6 Sol and Fable 5, Signaling End of Super Cheap Chinese AI / Ex-OpenAI CTO Mira Murati's Thinking Machine Releases First Model Inkling / Thinky Releases 975B Multimodal Open Model

Selected

9

Sources

21

2026-07-15

2026-07-15 AI Daily

Meet GPT-Red: An LLM Super-Hacker OpenAI Built to Make Its Models Safer / NVIDIA Introduces New Jetson Thor Computers to Advance Mainstream Robotics and Edge AI / Build a Multi-Camera 3D Tracking Application with NVIDIA DeepStream 9.1 / Katalyze AI Raises $10.5M to Build the Agentic Operating System for Pharmaceutical Companies

Selected

4

Sources

6

2026-07-12

2026-07-12 AI Daily

Claude Code 内置浏览器：AI 可读取、点击和输入外部网站 / AI代理在《杀戮尖塔2》中获胜：研究人员用结构化记忆替换增长聊天记录 / Meta取消Muse Image功能，该功能可在未经同意的情况下生成任何Instagram用户的AI照片

Selected

3

Sources

3

2026-07-11

2026-07-11 AI Daily

苹果起诉OpenAI，指控其通过挖角员工窃取商业机密 / 恐怖组织利用各大AI聊天机器人策划袭击和研发武器 / 中国Orca世界模型无需动作标签即可匹配专业机器人系统 / OpenAI GPT-5.6 Sol Ultra 据报在一小时内解决50年未解数学难题 / Meta的Muse Spark 1.1在编程任务上超越GLM-5.2且成本略低

Selected

5

Sources

4

2026-07-10

2026-07-10 AI Daily

GPT-5.6 Sol Ultra 产生环双覆盖猜想的证明 / NVIDIA BioNeMo 代理工具包加速端到端共折叠性能 / 基于JAX的LLM训练中通过主机卸载减少高带宽内存瓶颈 / OpenAI的GPT-5.6 Sol使用“相当不明确的提示”自主对较小的Luna模型进行后训练 / Slack 推出基于代理的端到端测试以提高UI测试自动化的韧性

Selected

7

Sources

6

2026-07-09

2026-07-09 AI Daily

OpenAI推出ChatGPT Work，可处理完整工作流程的智能助手 / Meta Muse Spark 1.1 API 以低价加剧 AI 模型价格战 / OpenAI GPT-5.6 系列模型 Sol、Terra、Luna 开始推出 / 本·伯南克被任命为Anthropic长期利益信托成员 / OpenAI AI在AtCoder全球总决赛中击败所有人类选手

Selected

10

Sources

13

2026-07-08

2026-07-08 AI Daily

OpenAI的GPT-5.6在美政府解除禁令后于周四发布 / GPT-Live 让与 AI 的对话如同真实交谈 / Muse Image技术令人印象深刻，但Meta使用Instagram照片引发争议 / OpenAI分析揭示SWE-Bench Pro编码基准测试问题 / Grok 4.5相比Fable 5和GPT 5.5价格低廉，基准测试差距可能不再重要

Selected

17

Sources

16