本文约4200字，推荐阅读时长15分钟，欢迎关注新财富公众号。 3月11日，Manus 官方正式宣布与阿里通义千问团队开展战略合作。由通义千问提供专属模型支持的 Manus 中文版本正在开发中… （由于没有邀请码测试 Ma n u s，本文侧重行业分析，体验归纳所选取的案例资料均来自于公开的测试回放） 1 创新的质疑 在 De e pSe e k 浪潮之后， "世界首个通用型AI Ag e n t "标签加持下的 Ma n u s ，注定要受到所有人的注目。 从自媒体争先恐后的"炸裂"、"上帝之手"般的吹捧，到"自嗨"、"饥饿营销"的千夫所指，Ma n u s在1 2个小时之内经历了从被吹上天到被锤到 地底的体验。 而这背后混杂的，是自媒体流量的争夺、投资人的FOMO（错失恐惧症）与普通用户对"一夜成神"故事的期待。 截至Ma n u s发布，海外AI巨头 An t h r o p i c 、Op e nAI 已经推出了Comp u t e r Us e 、Op e r a t o r这些垂类的AI通用Ag e n t，在代码垂类领域也有 De v i n AI 编程 Ag e n t 这样的产品，国内智谱也 ...

Investment Rating - The report maintains an "Overweight" rating for the industry, consistent with the previous rating [1]. Core Insights - The global first general-purpose AI agent, Manus, has entered internal testing, emphasizing independence and completeness in execution, accelerating the deployment of agents [5][7]. - Alibaba's DAMO Academy has launched the XuanTie RISC-V series chips, focusing on high performance and AI applications, with the highest performance processor C930 set to deliver in March [5][8]. - TSMC announced an additional investment of $100 billion in the U.S. to build five semiconductor factories, bringing its total investment to $165 billion [5][9]. Summary by Sections Weekly Overview - Manus, developed by the Chinese AI startup Monica, is designed to not only provide suggestions but also deliver complete task results, outperforming OpenAI's Deep Research in GAIA benchmark tests [7]. - The RISC-V architecture is gaining traction, with predictions that its market share will exceed 25% in various sectors, including consumer PCs and high-performance servers [8]. - TSMC's investment aims to alleviate geopolitical pressures and maintain a gross margin of over 53% [9]. Market Performance - The Hang Seng Index rose by 5.62%, and the Hang Seng Tech Index increased by 8.43% during the week from March 2 to March 8, 2025 [10]. AI Industry News - Tencent's Mixuan released an open-source video generation model, allowing users to create short videos from images [21]. - Alibaba introduced a new inference model, QwQ-32B, which outperforms existing benchmarks and is set to be open-sourced [22]. - OpenAI plans to launch a "PhD-level" AI agent with a subscription fee of up to $20,000 per month [21]. Investment Recommendations - Recommended stocks include NVIDIA (NVDA.US), TSMC (TSM.US), and Alibaba (9988.HK) among others, focusing on computing power, cloud services, and AI applications [23].

Group 1: Market Overview - The A-share market experienced a broad-based rally led by technology, with the Shanghai Composite Index rising by 1.55%, the Sci-Tech 50 up by 2.67%, the Shenzhen Component increasing by 2.19%, and the ChiNext Index gaining 1.61% during the week [1][9] - Daily trading volume reached a new high of 1.70 trillion yuan, indicating strong market activity concentrated in high-growth sectors such as AI, smart driving, and robotics [1][9] Group 2: Technology Sector Insights - The technology sector is driven by dual forces of AI computing power and data elements, with storage chips entering a price increase cycle, benefiting companies like Jiangbolong, which is positioned as a leader in the storage industry [2][21][22] - The connector market is also poised for growth, with companies like Hongrida benefiting from the demand for AIoT upgrades and increased orders from smart terminals [3][24] Group 3: High-end Manufacturing - The smart driving industry in China is rapidly developing, supported by policy initiatives and technological advancements, with expectations that by 2025, the penetration rate of high-level autonomous vehicles will exceed 70% [4][25] - The industry is witnessing a shift towards domestic chip production and the formation of a cloud-integrated ecosystem, with companies like BYD pushing advanced smart driving technology to more affordable models [4][25] Group 4: Mechanical and Industrial Automation - Increased infrastructure investment and domestic demand recovery are driving growth in the mechanical and industrial automation sectors, with a focus on intelligent and electric technologies reshaping the industry landscape [6][26] Group 5: Innovative Pharmaceuticals - The innovative drug sector is highlighted for its investment value, with Chinese pharmaceutical companies leveraging low-cost production and efficient R&D to enhance their global market share [7][27] - The sector is experiencing a shift towards original drug development, supported by policy reforms that accelerate the growth of biotech companies [7][27]

Summary of Conference Call on AI and Multi-Agent Systems Industry Overview - The conference focuses on the AI industry, particularly the development of chatbots and multi-agent systems, highlighting the transition from single agents to multi-agent systems as a significant trend in AI technology [2][3][7]. Key Points and Arguments 1. **Current State of AI Agents**: The development of AI agents is limited by model capabilities and engineering challenges. Despite high expectations for agents that can replace humans in complex tasks, no mature products have emerged yet [3][4]. 2. **Minus Product**: The Minus product is not an innovative model but offers a new approach to achieving multi-tasking capabilities within existing model limitations. It has sparked interest in the industry for practical applications of agents [4][5]. 3. **Multi-Agent Systems (MAS)**: MAS is a crucial direction in AI development, where multiple agents collaborate to compensate for individual limitations. This system enhances task automation and has shown promising results post-Minuse product launch [5][15]. 4. **Technological Breakthroughs in 2024**: Key advancements in AI technology include improvements in perception, definition, memory, planning, and action, laying the groundwork for more sophisticated multi-agent systems [6][10]. 5. **Action Mechanisms**: Significant breakthroughs in the action phase include virtual machine forms that address data source access issues and agent orchestration capabilities that assign tasks to the most suitable agents [9][10]. 6. **Progress in Large Models**: Large models have made notable progress in reasoning and action through methods like Chain of Thought (COT) and Reasoning + Acting, although human intervention remains common in enterprise applications [10][11]. 7. **Code Agent Development**: Code agents have matured, capable of automating various coding tasks and expanding their application scenarios beyond just code generation [11][12]. 8. **Data Access and Personalization**: The extent of data access is a critical factor in extending general scenarios, with companies like Apple and Tencent working on integrating personal behavior data for enhanced services [12][13]. 9. **MCP Protocol**: The MCP protocol is designed for cloud systems to ensure standardized information sharing and task collaboration among agents, which is vital for the development of multi-agent systems [13][14]. 10. **Enterprise Demand for MAS**: Companies have complex task orchestration needs, leading to significant interest in multi-agent architectures. Firms like Workday, ServiceNow, and Salesforce are exploring these systems to maximize their value [28][30]. Additional Important Insights - **Future of Multi-Agent Technology**: Multi-agent technology is expected to evolve from individual agents to a network, becoming a vital part of the next generation of the internet. This technology will play an increasingly important role in consumer devices [29][30]. - **Open Source Frameworks**: Various open-source multi-agent frameworks are emerging, providing users with customizable solutions to meet their specific needs [25][27]. - **Coordination Mechanisms**: Multi-agent systems utilize both static and dynamic coordination mechanisms, with dynamic approaches becoming more prevalent in current applications [23][24]. This summary encapsulates the key discussions and insights from the conference call, emphasizing the current state and future potential of AI and multi-agent systems in the industry.

图片来源： Unsplash 马克 ·扎克伯格今年正在提升 Meta 人工智能的语音功能，准备从这项快速发展技术中创收。 Meta 计划在其最新的开源大型语言模型 Llama 4 中引入改进的语音功能，预计将在未来几周内发布，据知情人士透露。 该公司押注未来的所谓 AI Agent 将是对话式而非文本主导的。 扎克伯格今年透露了，构建具备中级工程师编码和解决问题能力的 AI 工程助手的计划，他表示这一计划潜在拥有"非常庞大的市场"。 该集团首席产品官 Chris Cox 周三重点介绍了 Llama 4 的一些计划，称其将成为一个"全能模型"，语音将"原生支持对话，而不是将语音转换为文本，发送 文本到 LLM ，获取文本输出，再将其转换回语音"。 在摩根士丹利技术、媒体与电信会议上，他补充道： "我认为这对界面产品来说是一个巨大的突破，即你可以与互联网对话并询问任何问题。我觉得我们仍 在努力理解这一功能的强大之处。" Meta 也在讨论最新 Llama 模型应具备的输出限制以及是否降低这些限制，两位知情人士透露。 讨论之际，竞争对手纷纷推出新产品，新任 " AI 沙皇"、硅谷风险投资家 David Sacks ...

Core Viewpoint - The article discusses the launch and capabilities of OpenAI's "Deep Research," an AI agent that utilizes end-to-end reinforcement learning to enhance efficiency in complex information retrieval and reasoning tasks, significantly reducing the time required for knowledge work from hours to minutes [2][10][24]. Group 1: Product Overview - "Deep Research" is designed to retrieve information from multiple online sources and generate detailed reports, completing tasks in 5 to 30 minutes compared to hours for humans [6][10]. - The product is part of OpenAI's agent series, following the "Operator" agent, with plans for further expansions including a "Shards Seeker" agent [4][6]. - The development of "Deep Research" was inspired by breakthroughs in reasoning paradigms and aims to tackle complex tasks requiring extensive online research and creativity [7][10]. Group 2: Target Users and Applications - The primary users of "Deep Research" include knowledge workers in various fields such as market analysis, medical research, and personal planning [11][12]. - The product has shown significant utility in scientific research, helping users find relevant literature and data [12]. - It is also beneficial for personal tasks like shopping and travel planning, allowing users to save time and make informed decisions [18][19]. Group 3: Technical Mechanism and Innovations - "Deep Research" employs a fine-tuned version of OpenAI's advanced reasoning model, o3, specifically trained for complex web browsing and reasoning tasks [24][25]. - The model's ability to dynamically adjust its search strategies during information retrieval sets it apart from traditional search engines [25][26]. - The integration of a "Chain-of-Thought" summary allows users to understand the reasoning process behind the model's search strategies, enhancing transparency [25][26]. Group 4: Future Developments and Impact - Future plans for "Deep Research" include expanding its capabilities to access private data and improving its analytical functions for more complex tasks [37][38]. - The potential impact of "Deep Research" on various professions, particularly in consulting and healthcare, is significant, as it can drastically reduce the time spent on research tasks [39][40]. - The technology is expected to empower knowledge workers rather than replace them, enhancing their efficiency and decision-making capabilities [39][40].

Investment Rating - The report maintains a "Positive" investment rating for the media internet industry [4] Core Insights - The report emphasizes the potential of domestic open-source models to overcome computational and chip limitations, accelerating technological catch-up. Major internet companies are expected to undergo value reassessment in AI development, with a focus on companies like Tencent, Alibaba, Kuaishou, and Bilibili [4] - Tencent's AI products are rapidly iterating, with the launch of "Hunyuan Turbo S," which has doubled the articulation speed and reduced initial latency by 44%. This indicates a proactive shift in Tencent's AI business strategy [4] - The report highlights the importance of companies that embrace new technologies and possess data, user, and application advantages, suggesting a focus on companies like Kunlun Wanwei, Meitu, and others [5] Summary by Sections Industry Performance - The media sector ranked fifth among all industries with a 5.16% increase from March 3 to March 7, 2025 [10][14] AI Developments - The report discusses the launch of the AI Agent product Manus by the Monica.im team, which has achieved state-of-the-art performance in GAIA benchmark tests, surpassing similar models from OpenAI. Manus features autonomous planning and learning capabilities, indicating significant advancements in AI applications [6][23] Gaming Sector - The report suggests exploring the potential of AI in gaming, particularly in mobile internet games that have weaker commercialization. Companies like Tencent, NetEase, and others are recommended for their innovative approaches [7] Film and Television - The report notes that quality content like "Nezha: The Devil's Child" is expected to drive box office demand and suggests monitoring key film producers and cinema companies [7][39] - In the television sector, the report highlights the performance of popular series, with "Nan Hong" leading in viewership [44] Internet Companies - The report indicates that major internet companies like Tencent and Alibaba are showing resilience in their financial performance, with a focus on strategic adjustments in their organizational structures [8] Publishing and Media - The report advises monitoring state-owned publishing companies as they explore new business models and continue to push for industry consolidation [9] AI Investment Events - The report details ten AI investment events, with significant funding rounds, including a $2 billion investment in SSI, indicating strong market interest in AI and frontier technology [29][30]

Investment Rating - The report maintains an "Outperform" rating for the communication industry [5][6]. Core Insights - The government work report for 2025 emphasizes commercial aerospace as a leading emerging industry, shifting from "encouraging exploration" to "scale and standardized implementation," highlighting its core position in the national modernization industrial system [1][15]. - Manus, a Chinese startup, launched the world's first general-purpose AI agent, which can assist in various applications beyond stock analysis, marking a significant advancement in AI capabilities [2][12]. - The 2025 Mobile World Congress (MWC) showcased innovations in 5G-A technology from Huawei and ZTE, which are expected to accelerate technological progress and the adoption of AI terminals [3][32]. Summary by Sections Commercial Aerospace - The 2025 government work report prioritizes commercial aerospace, indicating a qualitative breakthrough in key technologies, market applications, and industrial ecology, which will inject new momentum into economic development [15][19]. - Various provinces, including Beijing, Shanghai, and Chongqing, are implementing special support policies for commercial aerospace, which is expected to enhance industrial cluster development [19][20]. AI Development - Manus's AI agent can perform tasks autonomously, utilizing a multi-agent architecture to simulate human workflows, and has achieved state-of-the-art performance in benchmark tests [2][10]. - The AI agent trend is becoming mainstream, with major tech companies accelerating their development, indicating a shift towards a more integrated AI application ecosystem [14][37]. 5G-A Innovations - Huawei's AI-Centric 5.5G solutions and ZTE's new 5G-A products were highlighted at MWC 2025, showcasing the integration of AI with 5G technology to enhance user experience and operational efficiency [23][24]. - The report notes that the communication sector index rose by 2.81%, outperforming the broader market, with specific sectors like military information technology and satellite internet showing strong performance [38][40]. Investment Recommendations - The report suggests focusing on companies benefiting from AI development, such as ZTE and Unisplendour, and highlights the high dividend potential of major telecom operators like China Mobile and China Telecom [43][44].

Investment Rating - The industry investment rating is "Positive" (maintained) [1] Core Viewpoints - The AI Agent sector is experiencing continuous catalysis, with applications reaching a "singularity moment" as demonstrated by the launch of Manus, an autonomous AI Agent product by Monica.im, which has achieved state-of-the-art performance in benchmark tests [4][11] - Broadcom's performance exceeded expectations due to strong demand for AI computing power, with significant investments from major tech companies in custom AI chips [5][12] - Alibaba plans to significantly increase investments in AI infrastructure and applications over the next three years, surpassing the total of the past decade [5][12] Market Review - During the week of March 3 to March 7, 2025, the CSI 300 index rose by 1.39%, while the computer index increased by 6.46% [3][10] Investment Recommendations - The report maintains a positive outlook on domestic AI industry trends, recommending companies such as Alibaba and DeepSeek, along with application-focused firms like Kingsoft Office and iFlytek [6][13] - In the computing power sector, recommended companies include Haiguang Information and Inspur Information, with beneficiaries like Cambrian and Unisplendour [6][13] - Cloud service providers recommended include Shenxinfu and Qcloud Technology, with additional recommendations for integrated machine solutions [6][13] Company Dynamics - True Vision announced winning a bid for the supply and installation of AV systems for the China Merchants Bank headquarters project, with a contract value of approximately 48.71 million yuan [15] - Zhongke Xingtu plans to establish a wholly-owned subsidiary in Shandong to support low-altitude economic development, with an investment of 100 million yuan [16] Industry News - Alibaba's international platform has integrated advanced AI models to enhance foreign trade processes, expected to attract more overseas customers [17] - ByteDance launched the domestic version of its AI programming tool Trae, aimed at improving developer efficiency [18] - Anthropic raised $3.5 billion in Series E funding, reaching a valuation of $61.5 billion, to enhance its AI systems and expand internationally [22]

Investment Rating - The report maintains an "Accumulate" rating for the industry [5] Core Insights - The latest Agent paradigm, Manus, showcases impressive multi-agent collaboration capabilities, solving complex tasks through independent thinking and system planning [1][14] - The rapid development of Agents highlights their application value, particularly in enhancing efficiency and decision-making across various industries [2][30] - The emergence of vertical Agents with high barriers to entry is expected to benefit software companies deeply, as they require industry-specific knowledge and tools [3][35] - AI advancements are transforming the domestic IT landscape, with significant potential for valuation recovery compared to the US market [4][40] Summary by Sections Section 1: Manus and Multi-Agent Collaboration - The Manus product can handle diverse tasks autonomously, utilizing various tools and executing code within a virtual environment [1][14] - Manus outperformed OpenAI's DeepResearch in the GAIA benchmark, achieving state-of-the-art performance across all difficulty levels [15][17] Section 2: Rapid Development of Agents - Key components of Agents, including planning, memory, and tool usage, are advancing, enhancing their capabilities [2][24] - The integration of AI Agents into business processes is expected to redefine core competencies and improve operational efficiency [30][31] Section 3: Vertical Agents and Software Companies - Vertical Agents require specialized software teams to develop, creating a significant opportunity for companies with industry expertise [3][36] - The development of AI Agents faces challenges due to the need for deep integration with specific industry knowledge and real-time data processing [36][37] Section 4: Transformation of the Domestic IT Landscape - AI advancements are leading to a fundamental change in the domestic IT sector, with potential for significant growth in software companies and related infrastructure [4][40] - The valuation of domestic companies remains relatively low compared to their US counterparts, indicating substantial room for growth as AI applications deepen [41][42] Section 5: Investment Recommendations - The report suggests focusing on various sectors, including IAAS, domestic computing power, SAAS, and AI-related companies, highlighting specific firms for potential investment [45][46]