Intelligence Archive
Historical Record / Total 214 Articles
March 2026
Google TurboQuant Algorithm Reshapes AI Efficiency, Sending Storage Chip Stocks Plummeting
Google researchers have introduced a new algorithm called TurboQuant, which drastically reduces AI memory usage by 1/6 and boosts inference speeds by 8x. The technology has caused a market crash in the storage chip sector, with major companies seeing significant stock drops.
GLM-4.7-FlashMoonshot AI Founder Yang Zhilin Cites Algorithm Scaling for Stable Pricing, Outlines AI Agent Adoption Roadmap
At the 2026 Zhongguancun Forum, Moonshot AI founder Yang Zhilin confirmed the company has not raised prices thanks to algorithmic cost reductions, while predicting phased AI agent adoption from early adopters to universal productivity penetration.
Westlake University Unveils 'Ultimate Avatar' Humanoid Robot Titan o1 with Real-Time Human Motion Replication
Westlake University spin-off西湖机器人公司 has launched the Titan o1 humanoid robot, featuring the world's first General Action Expert (GAE) system that enables real-time motion replication with one-to-many remote control capabilities, claiming a six-month technological lead over international competitors.
SenseTime Reports Record Revenue of Over 5 Billion RMB, Eyes Q2 2026 Model Launch
Chinese AI giant SenseTime has announced a record-breaking financial year for 2025, with total revenue surging 33% to exceed 5 billion RMB. The company also revealed plans to launch a second-generation NEO architecture model in the second quarter of 2026, building upon its earlier open-source Native VLM.
GLM-4.7-FlashMiniMax Launches World's First Subscription Plan for Full-Modal AI Models
MiniMax has upgraded its Coding Plan to the new Token Plan, the world's first subscription service supporting full-modal AI models. The plan integrates programming, video, speech, music, and image generation capabilities into a single package with included usage allowances.
Alibaba's PrismAudio Framework Teaches AI to 'Think Before It Makes Sound'
Alibaba's Tongyi Lab has unveiled PrismAudio, a novel 'Video-to-Audio' framework that uses a 'chain-of-thought' reinforcement learning approach to generate perfectly synchronized environmental sound effects like footsteps and rain from silent video. It features a unique four-instructor evaluation system to ensure semantic, timing, aesthetic, and spatial accuracy.
Alibaba International Debuts Enterprise AI Agent Accio Work, Promises 30-Minute Autonomous Store Creation
Alibaba International has launched Accio Work, its first enterprise-grade AI agent for overseas markets, featuring autonomous capabilities to build e-commerce stores, manage supply chains, and execute marketing campaigns with minimal human intervention.
iPhone 17 Pro Runs 400-Billion-Parameter AI Model, But at a Crawl
A developer has successfully run a 400-billion-parameter large language model on an iPhone 17 Pro using a technique called Flash-MoE, which streams data from the device's SSD to the GPU. However, the achievement comes with a major caveat: an excruciatingly slow token generation speed of just 0.6 tokens per second.
China's Daily Token Usage Surpasses 140 Trillion, Marking a 1,000-Fold Increase in Two Years
At the China Development Forum 2026, Liu Liehong, head of the National Data Bureau, revealed that China's daily Token usage exceeded 140 trillion in March, representing a growth of over 1,000 times in two years. The bureau designated 2026 as the 'Data Value Release Year' and announced plans to enhance computing-electricity synergy.
GLM-5Baidu Launches DuMate: First Domestic Enterprise-Grade 'Full-Strength' OpenClaw
Baidu's intelligent cloud platform DuMate has officially gone live, featuring local deployment, robust security isolation, and native office software support as the first domestic enterprise-grade 'full-strength' OpenClaw.
GLM-4.7-Flash智元酷拓 D1 MAX 四足机器人发布:30 公斤载重,30 公里续航
智元机器人发布了新款行业四足机器人酷拓 D1 MAX,具备30公斤载重、30公里续航、IP67防护等级及8米/秒的速度,适用于安防、巡检、物流等多个领域。
Alibaba Qianwen Integrates Taxi-Hailing with Natural Language Control
Alibaba's Qianwen AI assistant now enables users to book rides, select vehicle types, add waypoints, and schedule trips using conversational commands, with support for nuanced preferences like driving stability and vehicle air quality.