Group 1 - The concept of "embodied intelligence" has gained significant attention, with practical demonstrations at the 2025 Zhongguancun Forum showcasing robots like "Tiangong" that can navigate complex terrains using advanced motion control capabilities [1][2] - The "Tiangong 2.0" robot features enhanced capabilities, including a self-developed dual battery quick-swap system and industrial-grade upper limb load capacity, allowing it to perform tasks with high precision [2] - The retail industry is beginning to experiment with embodied intelligence, as demonstrated by the Galbot robot, which utilizes advanced technologies for fully automated operations, improving inventory management and supply chain efficiency [2] Group 2 - Embodied intelligence integrates artificial intelligence into physical entities, enabling them to perceive, learn from their environment, and interact dynamically [1] - The Beijing Embodied Intelligence Technology Innovation and Industry Cultivation Action Plan (2025-2027) aims to achieve breakthroughs in over a hundred key technologies and promote the deployment of thousands of embodied robots, fostering a trillion-yuan industry cluster [3] - Various innovative applications of embodied intelligence were showcased, including a mechanical arm capable of performing complex tasks without prior training and a highly dexterous robotic hand developed by a Tsinghua University incubator team [3]

【环球时报报道 记者 李迅典】3月27日至31日，以"新质生产力与全球科技合作"为主题的2025中关村论 坛年会在北京举行。本次论坛期间，中国数十家顶尖科技公司的近百台人形机器在迎宾、交流、表演等 场景中为与会嘉宾提供服务。《环球时报》记者在论坛年会期间采访了多名业内专家，共同探讨人形机 器人在未来世界中的无限可能。 AI+机器人 韩国大使馆公使衔参赞李镇守在论坛现场表示，中国大模型DeepSeek非常出圈，中国AI技术已经达到 全球领先水准，具备非常强的竞争力，这种AI领域的革新也是与机器人产业紧密联结在一起，互相促 进了各自的发展。纵观中国的机器人产业和政府政策，他表示，韩国可以借鉴很多东西，在中国春节晚 会上跳舞的机器人也对全世界带来了很大的震撼。李镇守说，中国也在大力推进机器人国产化，2022年 中国工业机器人销量中的52.5%来自中国本土制造。 人形机器人的发展离不开关键技术的持续突破，在人工智能、机器视觉、语音识别等技术领域的不断增 强下，人形机器人已具备自主学习、自主决策，更好地适应各种复杂环境的能力。意大利机器人国家能 力中心Artes4.0首席科学家、意大利圣安娜高等研究院生物机器人研究所所 ...

编者按：人形机器人的热度，让那些幕后的企业创始人不自觉地被推向台前。这些科技新贵受到了前所未有的关注。在2025中关村论坛年会期间，北京商报 记者专访了三位来自一线的当局者，一边是行业之内，围绕产业拐点、商业化逻辑、量产时间表等，试图去给无数好奇的旁观者解惑；一边是行业之外，聚 焦科技、流量、创业，想看看科技与大众现象级"亲密互动"到底意味着什么，会改变什么，又会创造什么。 冷晓琨（右）接受北京商报记者专访（张笑嫣/摄） 入行9年，2025年是乐聚通研（北京）机器人技术有限公司（以下简称"乐聚机器人"） 创始人冷晓琨最忙的一年，日程以半小时计。 3月26日，2025中关村论坛年会开幕的前一天下午，他从山东的工厂赶到北京，3月29日下午结束未来人工智能先锋论坛圆桌讨论后，他又直奔北京首都国际 机场，"继续回归产业和研发，闭关修炼"。 这期间，乐聚人形机器人们也没闲着，在中关村论坛年会这个国际化舞台，它们既参与开幕式表演，又提供服务。"我们的核心目标已经基本达成，让大家 接触了人形机器人，我们的产品也得到了历练"，每次说到自家机器人，他总是笑意盈盈。 从早期只有个位数的同行，到现在几百家共舞，乐聚机器人的策略没有 ...

Core Insights - The article discusses the rapid development of emerging industries, particularly in artificial intelligence and humanoid robots, and the role of financial capital in supporting the growth of technology companies [2][6] - It highlights the importance of financial asset investment companies (AICs) in promoting equity investment to support technological innovation and the government's emphasis on developing patient capital [1][2] Group 1: Characteristics of Technology Company Growth - Technology companies typically experience five growth stages: emergence, expansion, trough, recovery, and maturity [6] - The growth of these companies is characterized by significant fluctuations, rapid transitions, and multiple development nodes [6][7] Group 2: Role of Financial Capital - Financial capital can play a counter-cyclical role to mitigate the impact of market volatility on technology companies [7] - It can provide comprehensive services tailored to the changing financial needs of technology companies at different stages of development [7][8] - Financial capital can actively participate in corporate governance to ensure balanced interests among stakeholders and promote long-term sustainable development [8] Group 3: Investment Criteria for Financial Institutions - The alignment with national strategic goals is crucial for investment decisions, ensuring that companies contribute to broader economic and social benefits [9][10] - The professional capabilities and dedication of the core team are essential factors in evaluating potential investments [9][10] - A robust corporate governance structure is necessary for effective execution of business plans and risk management [10] - Companies must possess distinctive resources and competitive advantages to thrive in a competitive market [10]

Group 1 - The "Civil Wisdom, Shared Sci-Tech Innovation" forum was held on March 30, 2025, focusing on how social organizations can support technological innovation development, particularly in the context of the Belt and Road Initiative [1] - Technological innovation cooperation is a key component of the Belt and Road Initiative, with social organizations, especially technology-oriented ones, playing an increasingly important role in this process [1][3] - The China Association for International Exchange of Personnel (CAIEP) emphasizes the unique advantages of social organizations in bridging differences and fostering consensus in global technology governance [3] Group 2 - Beijing is recognized as China's international technology innovation center, housing over 92 universities and more than 1,000 research institutions, with over 550,000 researchers [4] - As of the end of 2024, there are over 850,000 technology-based enterprises in Beijing, with approximately 110,000 new technology enterprises established annually, indicating a daily creation of over 300 technology firms [4] - The establishment of the Zhongguancun Global High-end Think Tank Alliance in September 2020 aims to enhance international cooperation through multilateral think tank exchanges [4][5] Group 3 - The Zhongguancun Global High-end Think Tank Alliance has expanded to include 95 high-end think tanks from 16 countries and regions, covering various types of think tanks [5] - The alliance has also established a "Belt and Road" International Technology Cooperation Training Center, which has trained over ten thousand technology personnel from China and abroad [5] - The Zhongguancun Social Organization Federation plans to expand its international cooperation network by adding 1-2 overseas liaison offices in emerging markets along the Belt and Road in the next three years [5]

Group 1 - As of the end of 2024, the number of technology-based enterprises in Beijing has exceeded 850,000, with 110,000 new technology enterprises established last year, averaging over 300 new tech companies formed daily [1] - Beijing is recognized as a global technology innovation center, housing 92 higher education institutions and over 1,000 research institutes, with more than 550,000 research personnel [3] - The city has maintained its position as the top global city for technology and research for eight consecutive years, leading in national laboratory numbers, basic research talent density, and R&D investment intensity [3] Group 2 - The forum highlighted the importance of social organizations in global technology governance, emphasizing their unique advantages in building cross-national dialogue platforms and promoting collaborative technology rule-making [1] - Legal frameworks are essential for international technology cooperation, especially in the context of increasing environmental degradation, trade wars, and technological barriers [4] - Since 2019, the Zhongguancun Forum has been hosting parallel forums focusing on technology innovation from a civil perspective, attracting over 120 participants and reaching more than 30,000 online viewers [5]

当你还在惊叹科幻电影中机器人与人类协同工作的场景时，具身智能正悄然将这一想象变为现实。 在2025中关村论坛年会上，具身智能成为科技领域的热门话题，不管是智慧安防还是物流配送，农业搭棚抑或搬运服务，甚至于养老陪护、科研教育等领 域，均有具身智能的触达。 从想象走进现实，具身智能崭露头角，尽管仍处早期，但业内普遍认为，给具身智能几年的成长时间，它将在更多领域和开放性场景爆发巨大潜力。 从某国企正在招聘的高级智能操作系统架构师岗位信息来看，其月薪在8万—12万元，负责装备智能操作系统理论及技术创新研究、组织装备智能操作系统 研发等，要求学历在硕士以上，年龄不超过35岁，并有知名高校副教授或以上职称，或者在操作系统相关企业担任高级技术岗位。 "别看高薪，但真正匹配的人才也难寻，很多岗位我们招聘了很久，收到的简历不少，但真正符合要求的专业人才寥寥无几。"前述具身智能公司HR无奈地 表示。 高薪背后，对人才的要求确实不低。北京商报记者注意到，另一家正在招聘AI人工智能高管的公司正在孵化集团AI项目，要求该岗位负责AI人工智能技术 应用规划工作，并负责AI人工智能团队建设工作，月薪在10万—12万元，但在岗位要求上，需要 ...

Group 1 - The core theme of the forum is the rapid integration of artificial intelligence (AI) into everyday life, transforming it from a futuristic concept to a present reality [1][2] - AI is becoming a crucial variable in various sectors such as manufacturing, automotive, and robotics, enhancing productivity and innovation [2][3] - The forum showcased advancements in AI applications, including a robot that can make coffee and a financial AI fraud prevention product that can intercept deepfake scams in milliseconds [1][2] Group 2 - The concept of "new quality productivity" was a focal point, with various technologies and innovations being presented that exemplify its development [3] - The establishment of research institutions and the introduction of new biomanufacturing technologies indicate a significant shift towards sustainable and innovative production methods [3] - The forum emphasized cross-industry collaboration, highlighting the potential for breakthroughs when different technologies and sectors intersect [3] Group 3 - The forum served as a global platform for innovation, featuring participants from 34 countries and showcasing over 70 projects, demonstrating the international nature of technological advancement [4][5] - The growth of patent applications in generative AI, which increased by over 800% since 2017, underscores the competitive landscape in digital technology innovation [4] - Major scientific facilities in China are now open to global researchers, providing extensive resources for various fields, which enhances international collaboration in scientific research [5]

Core Insights - The topic of embodied intelligence is a major focus at the 2025 Zhongguancun Forum, with the introduction of the RoboOS framework and the open-source RoboBrain model [1][3] - Multi-modal large model technology is expected to enhance the intelligence of robots, allowing them to better understand and interact with the physical world [2][3] Group 1: Multi-modal Large Models - Multi-modal large models enable AI to perceive and understand the world through various data types, such as medical imaging and sensor data, facilitating the transition from digital to physical environments [2] - The performance improvement of large language models has slowed due to the exhaustion of available internet text data, necessitating the integration of multi-modal capabilities [2] Group 2: RoboBrain and RoboOS - RoboBrain and RoboOS are designed to support cross-scenario, multi-task deployment and collaboration among different types of robots, enhancing their general intelligence [3] - RoboBrain can interpret human commands and visual inputs to generate actionable plans based on real-time feedback, supporting various robotic configurations [3] Group 3: Industry Development and Challenges - The open-source approach is seen as a key driver for rapid development in the AI industry, allowing for collaboration among hardware, model, and application vendors [4] - Despite the potential of humanoid robots, there are significant challenges in their industrial application, with many still in the early stages of development [5] - The realization of Artificial General Intelligence (AGI) is projected to take an additional 5-10 years, influenced by advancements in embodiment capabilities and data accumulation [5]

Core Insights - The cost of inference for large models is decreasing rapidly, at a rate of tenfold per year, which is crucial for the explosion of AI-First applications [2] - The year 2025 is anticipated to be a pivotal year for the explosion of AI-First applications and the practical implementation of large models [2] - The growth of pre-training scaling laws is slowing down due to data volume and computational limitations, but a new "slow thinking" scaling law is emerging that focuses on longer inference times for better results [2][3] Industry Developments - The performance growth of models under the slow thinking scaling law is accelerating, indicating significant potential for further advancements [3] - The Chinese market is experiencing a "DeepSeek Moment," which is facilitating the awakening of AI-First applications, overcoming previous barriers [3] - The company has made strategic adjustments to fully embrace DeepSeek, focusing on transforming high-quality base models into enterprise-level customized solutions, akin to creating a Windows system for the AI 2.0 era [3]