Group 1 - Musk's AI company xAI is collaborating with Valor Equity Partners to raise up to $12 billion for acquiring advanced Nvidia chips and building large data centers to support its AI chatbot Grok [1] - Shanghai AI Laboratory has launched a multi-agent framework called Intern·Agent, which enhances research efficiency by enabling AI to autonomously explore and conduct experiments across 12 scientific fields [1] - Alibaba is set to release its first self-developed AI glasses, featuring basic functionalities like voice assistance, music playback, and real-time translation, marking a significant step in its AI to C strategy [1] - Tongyi Qianwen has officially open-sourced its latest AI programming model Qwen3-Coder, which boasts 480 billion parameters and supports a context length of up to 1 million [1] - OpenAI's CEO stated that AI has the capability to bypass most authentication methods, indicating a potential major fraud crisis in the financial sector [1]

此外，上海AI实验室推出多智能体框架Intern·Agent，并探索以此为核心构建科学发现系统。通过调动 多个专职智能体，Intern·Agent使AI获得类似人类科学家的自主探索能力，可实现"想法生成-方案设计- 代码实现-实验验证"全流程闭环。 兴业证券表示，大模型技术进入快速迭代期，AI产业迎来密集催化，AI应用规模商业化落地置信度增 强，看好端侧AI落地场景。 2、政策层面，昨日（7月22日）是科创板开市六周年！2025陆家嘴论坛上，高层提出在科创板设置科创 成长层，重点服务人工智能等前沿科技领域，技术有较大突破、商业前景广阔、研发投入大的科技企 业。此外，工信部发布会表态，近期会出台多个行业数字化转型方案，组织"人工智能+软件"行动，并 推动未来产品创新发展。 3、产业层面，景气度向上，AI应用端加速渗透，大模型的技术持续突破：①国产AI公司月之暗面研发 的Kimi K2强势登顶全球开源模型榜首！②OpenAI推出全新智能体产品"ChatGPT Agent"，这一新型AI Agent具备自主思考与行动能力。 此前七连涨的科创人工智能ETF华宝（589520）在昨日暂歇后，今日（7月23日）再度雄起， ...

Core Viewpoint - Shanghai Artificial Intelligence Laboratory has launched a multi-agent framework called Intern·Agent, aimed at enhancing scientific discovery processes through collaborative AI agents [1] Group 1: Framework Features - Intern·Agent enables multiple specialized AI agents to work collaboratively, mimicking the autonomous exploration capabilities of human scientists [1] - The framework facilitates a complete closed-loop process from idea generation to experimental validation, including problem analysis, literature review, research reflection, and experimentation [1] Group 2: Implementation and Impact - Intern·Agent has been tested in 12 research fields, significantly improving research efficiency [1]