The AI community has been quite lively in the past 24 hours, and there are several messages worth mentioning: 1. NVIDIA has released Nemotron 3 Nano Omni - an open-source multimodal model that integrates visual, speech, and text, claiming to have increased performance by 9 times and directly topped 6 charts. Open source, lightweight, 306B parameters, suitable for local deployment. The direction is clear: in the future, AI agents will no longer need to piece together different models to handle different modalities. OpenAI and Microsoft officially end their exclusive partnership. OpenAI's model is now available on AWS Bedrock, including the Codex programming agent. Microsoft will no longer contribute revenue sharing to OpenAI, but will retain IP authorization until 2032. The two sides have changed from a 'bound marriage' to an 'open relationship'. At the same time, it was reported that OpenAI's revenue did not meet its internal targets, putting considerable pressure. Time magazine has released the top 10 most influential AI companies of 2026. ByteDance, Zhipu, and Alibaba are three Chinese companies on the list. This ranking does not look at parameter rankings, but rather at the impact on the entire industry path and society. On the domestic side, the Ministry of Industry and Information Technology and the National Data Administration jointly issued a notice on the "Modular Resonance" action, aiming to promote the implementation of AI scenarios in 20 industries - paving the way for intelligent agent factories at the policy level. 5. Shenzhen independent variable robot released WALL-B embodied intelligent model. Based on the "World Unified Model Architecture", vision, language and action are all jointly trained in a network. Next month, we will bring the new generation of robots into households. Another interesting thing is that the EU requires Google to open up its AI capabilities on Android, but Google directly says this is an "unreasonable intervention". The friction between Silicon Valley and Brussels over AI regulation is heating up. The pace of the AI industry is becoming increasingly intense - open source models, business models, policies, geopolitics, and several lines are accelerating simultaneously.