02月24日AI资讯 | DeepMind CEO预测AGI 2030年实现;字节Seed用化学思想分析AI推理;千寻智能融资20亿引爆具身智能;Anthropic指控中国大模型侵权;OpenClaw安全事件引发行业关注DeepMind CEO Demis Hassabis predicts AGI will be achieved by 2030, emphasizing the need for international cooperation on safety. ByteSeed uses chemical concepts to analyze AI reasoning, improving model stability. Qianxun Intelligent raises 2 billion in funding, becoming a leader in embodied intelligence. Anthropic accuses Chinese companies of infringing on its Claude model, while OpenClaw's security issues highlight AI's risks. Predictions suggest AI may soon produce top-tier physics papers, and new methods are proposed to measure AI's economic contributions.