Investment Rating - The report maintains a "Positive" investment rating for the automotive and automotive parts industry [11] Core Insights - The report highlights a significant leap in the development of general artificial intelligence large models, with continuous breakthroughs in model scale, training paradigms, and reasoning capabilities, establishing a solid technological foundation for various AI applications. Intelligent driving, being an application of "physical AI," is evolving towards large models, marking a new phase of intelligent emergence [3][6] Summary by Sections Introduction: AI Empowerment, Intelligent Driving Enters the Large Model Era - The report discusses the rapid development of general artificial intelligence large models, emphasizing their role in enhancing intelligent driving through technological iterations [6][19] Emergence of General Large Model Capabilities - The AI large model era is characterized by the use of the Transformer architecture, exponential increases in computing power, and the accumulation of vast multimodal data, leading to critical breakthroughs in AI applications [7][21] Progression of Intelligent Driving Large Models - Intelligent driving has transitioned from rule-based models to end-to-end large models, gradually evolving towards VLA (Vision-Language-Action) and world models, enhancing deep reasoning and decision-making capabilities [8][50] Investment Recommendations - The report suggests that the continuous emergence of AI large model capabilities will accelerate the commercialization of high-level intelligent driving. Key recommendations include companies like XPeng Motors, BYD, and Geely in the vehicle sector, and Top Group and Bertelson in the parts sector [9]

前Meta公司AI负责人杨立昆（Yann LeCun）也创立了一家"世界模型"初创公司AMI Labs。杨立昆认 为，目前市面上的机器人缺乏对物理世界的基础模型理解，教会AI 理解物理世界的能力至关重要。 谷歌DeepMind也在积极开发世界模型，该公司的Genie模型可以生成和模拟三维环境。 当地时间2月18日，人工智能先驱之一的华人科学家李飞飞创立的AI初创公司World Labs宣布在新一轮 融资中筹集了10亿美元，用于推动所谓的"世界模型"的研发。 在这轮融资中，软件公司欧特克向World Labs投资了2亿美元，其他投资方还包括英伟达以及AMD。尽 管World Labs没有透露估值，但市场预估该公司的估值约在50亿美元。 去年年底，World Labs推出了首个空间智能产品Marble。该公司将Marble描述为一种可以根据图像或文 本提示生成三维世界的基础模型，通过处理来自周围物理环境的视觉数据，开发出高级推理的能力。 借助这笔新融资，World Labs将把重点放在提升机器人和科学发现等领域的应用能力。李飞飞表示，空 间智能模型在未来可以用于增强现实和虚拟现实或机器人。 World Labs" ...

Core Insights - The AI landscape is rapidly evolving, with significant advancements in large models, intelligent agents, computing power, data centers, and application ecosystems, leading to a competitive environment where companies like OpenAI, Anthropic, and Google are vying for market leadership [2] - The focus for 2026 will shift from merely enhancing model capabilities to a comprehensive evaluation of cognitive direction, implementation pathways, and organizational capabilities [2] - The transformation of AI from a tool to a collaborative partner is crucial, emphasizing the need for governance and the establishment of boundaries for AI decision-making [4][5] Group 1: AI Evolution and Organizational Change - The current AI revolution differs from past technological waves as it transcends mere tool iteration, fundamentally altering the boundaries of productivity factors [3] - The concept of "Superintelligence" emerges as a new organizational form where companies operate as continuously learning systems, integrating machines, humans, and organizations [3] - The transition to a "human-centered" approach in AI is essential, where machines evolve to become proactive agents capable of executing complex tasks [4] Group 2: Governance and Human-Centric Focus - As AI becomes more autonomous, the governance of AI systems will be critical, necessitating clear boundaries for AI actions and human oversight [5] - The shift from mechanical tasks to human-centric roles will highlight the importance of human capabilities such as vision, judgment, and creativity in navigating AI's influence [6] - Companies must adapt to a new organizational structure that fosters human-machine collaboration, moving away from traditional hierarchical models [6] Group 3: Practical Application and Infrastructure Investment - The focus on "real-world" applications of AI will be paramount in 2026, requiring companies to demonstrate scalable and replicable AI solutions that yield tangible business results [7] - Investment in AI infrastructure is expected to surge, with major tech companies forecasting significant capital expenditures to support AI deployment [8][9] - The evolution of AI models will expand beyond language models to include architectures that can understand physical laws and facilitate real-world applications [8] Group 4: Future Outlook and Strategic Priorities - Despite skepticism regarding investment returns, leading companies are prioritizing AI as a key strategic initiative to drive measurable value [11] - The dual focus on enhancing human capabilities alongside AI deployment will be essential for maximizing productivity and ensuring effective organizational transformation [11]

Core Viewpoint - The announcement by Hang Seng Index Company regarding the inclusion of Pony.ai (02026) in the Hang Seng Composite Index is a significant development, expected to enhance the company's visibility and investment potential in the market [1] Group 1: Index Inclusion - Pony.ai is set to be included in the Hang Seng Composite Index as of March 6, 2026, with the changes taking effect on March 9, 2026 [1] - The inclusion will lead to adjustments in the eligible stocks for the Hong Kong Stock Connect, as per the regulations of the Shanghai and Shenzhen stock exchanges [1] - According to CICC's research report, Pony.ai meets various criteria such as market capitalization, liquidity, and listing duration, making it a candidate for inclusion in the Hong Kong Stock Connect [1] Group 2: Strategic Partnership - On February 6, Pony.ai, known as the "first global Robotaxi stock," announced a strategic partnership with Moore Threads (688795.SH), a leading domestic full-function GPU company [1] - The collaboration will focus on the implementation and scaling of Level 4 autonomous driving technology, emphasizing deep cooperation on training and optimization of Pony.ai's core technologies, including the world model and virtual driver system [1] - The partnership aims to leverage safe and reliable AI computing power to facilitate the iteration and commercial deployment of autonomous driving technologies [1]

Core Insights - The emphasis on technological self-reliance and innovation is crucial for building a modern socialist country, as highlighted by President Xi Jinping during his visit to the National Innovation Park in Beijing [1][2] Group 1: Government Support and Policies - The government provides substantial policy benefits, such as tax reductions, which lower innovation costs and operational burdens for companies like VAST, allowing more funds to be directed towards research and development [1] - The presence of "patient capital" in Beijing, particularly from government-guided funds, encourages long-term investments in high-cost areas like 3D modeling and world models [1] Group 2: Talent and Innovation Capabilities - Beijing is recognized as a hub for high-level talent, with top universities and research institutions attracting the best algorithm experts, which is beneficial for companies like VAST [2] - The leadership team at VAST includes highly qualified individuals, such as the CTO and Chief Scientist, both of whom graduated from Tsinghua University, showcasing the strong talent pool available [2] Group 3: Application and Validation Scenarios - The city offers abundant scenarios for concept validation and application in fields like smart manufacturing, virtual reality, and embodied intelligence, providing a robust environment for transforming laboratory algorithms into scalable applications [2] - The integration of these advantages is seen as a solid pathway for Beijing to accelerate its development as an international innovation center [2] Group 4: Company Vision and Goals - VAST aims to contribute to the establishment of a technology powerhouse in China by focusing on independent research and technological breakthroughs that align AI technology with a strong industrial base [2] - The company is committed to the deep integration of technological and industrial innovation, emphasizing the importance of self-reliance in core technologies and the empowerment of industries through innovative outcomes [2]

图片来源： Runway 知情人士 透露， AI 视频生成初创公司 Runway 已完成 3.15 亿美元 E 轮融资，公司估值飙升至 53 亿美元，较之前水平近乎翻倍。 公司在其宣布融资的博客中表示，新资金将使 Runway 能够 " 预训练下一代世界模型，并将其引入新产品和行业 " 。 世界模型是一种能够构建环 境内部表征的人工智能系统，从而能够对未来事件进行规划，许多顶尖学者认为这类模型对突破大语言模型的局限至关重要。 据公司发言人透露，展望未来， Runway 计划运用新资金将其约 140 人的团队在研发、工程和市场拓展等岗位进行快速扩容。 本轮融资由 General Atlantic 领投，参投方包括英伟达、富达管理与研究公司、 AllianceBernstein 、 Adobe Ventures 、未来资产、 Emphatic Capital 、 Felicis 、 Premji 以及 AMD Ventures 。 参考资料： https://techcrunch.com/2026/02/10/ai-video-startup-runway-raises-315m-at-5-3b-valuatio ...

Core Insights - The development of intelligent driving is not a linear technological curve but a result of the interplay between various technical paradigms, engineering constraints, and real-world scenarios [1] - As the industry moves beyond the proof-of-concept stage, single technical terms can no longer explain the real differences in capabilities [2] - Factors such as computing power, data quality, system architecture, and engineering stability are determining the upper and lower limits of intelligent driving [3] Group 1: Evolution of Learning Techniques - Recent discussions in intelligent driving technology reveal a trend where various paths, such as end-to-end, VLA, and world models, converge on the concept of reinforcement learning [5] - Reinforcement learning is transitioning from a "technical option" to a "mandatory option" in the industry [7] - The emergence of products like AlphaGo and ChatGPT has highlighted the effectiveness of allowing AI to learn through trial and error as the fastest evolutionary method [8][9] Group 2: Learning Methodologies - Understanding reinforcement learning requires a grasp of imitation learning, which was previously favored in intelligent driving [11] - Imitation learning allows AI to learn from human driving data but has limitations, such as inheriting bad habits and struggling with unfamiliar situations [14][16] - Reinforcement learning, as demonstrated by AlphaGo, allows AI to explore new strategies through self-play, leading to superior performance beyond human intuition [17] Group 3: Reinforcement Learning Mechanisms - Reinforcement learning operates on a trial-and-error basis, where the model learns to drive well through a cycle of feedback [26] - The design of reward functions is crucial, as it translates driving performance into quantifiable scores [30] - Balancing conflicting objectives, such as safety versus efficiency, is essential in reward function design [32] Group 4: World Models and Advanced Learning - The integration of world models with reinforcement learning enhances the training environment, allowing AI to simulate real-world scenarios [42][49] - High-fidelity virtual environments enable AI to consider long-term consequences of actions, improving decision-making [50] - The coupling of world models and reinforcement learning creates a feedback loop that accelerates model iteration and performance [52] Group 5: Industry Trends and Future Directions - The importance of data is being redefined, with a shift towards the ability to model the world rather than just relying on raw data [56] - Companies are focusing on enhancing the "modeling capacity" of their systems, which is crucial for intelligent driving [60] - The evolution of intelligent driving systems is moving towards a stage where AI can independently understand environments and refine strategies, marking a significant advancement in the industry [62]

Core Viewpoint - The humanoid robot industry is shifting focus from entertainment to practical applications, with a significant increase in production expected in the coming years, driven by the need for robots to demonstrate real-world value rather than just perform on stage [1][3][4]. Industry Trends - The humanoid robot market is projected to see a shipment increase of over 650% in 2025, reaching approximately 18,000 units, and is expected to rise to 62,500 units in 2026 [3]. - The industry is moving past a phase of "wild growth," where mere performance was sufficient for sales, to a more mature phase where practical applications and real-world scenarios are essential for success [4][14]. Technology Development - Three main technical routes are emerging in the humanoid robot sector: VLA (Visual Language Action) model, world model, and layered decision-making with hardware-software collaboration, each with distinct advantages and challenges [7][9][10]. - The VLA model aims for general intelligence, relying on vast data for training, while the world model focuses on simulating physical environments to predict actions [9][10]. - The layered decision-making approach breaks down complex tasks into manageable components, enhancing reliability and efficiency in real-world applications [9][10]. Market Demand - There is a growing demand for robots that can operate in specific scenarios, such as factories and logistics, where they can perform tasks like assembly and packaging, thus providing tangible economic value [13][14]. - Users are increasingly looking for robots that can reduce production costs and alleviate humans from repetitive or hazardous tasks, indicating a shift towards practical applications [13][14]. Investment Focus - Investors are prioritizing companies that can demonstrate viable application scenarios and tangible products, moving away from those that lack a clear path to market [4][5]. - The consensus among investors is that companies must integrate their technology with real-world applications to avoid being sidelined in a competitive landscape [4][5]. Future Outlook - The next 3 to 5 years are critical for the practical deployment of humanoid robots, with expectations that they will increasingly complement human labor rather than replace it [13][19]. - The industry is expected to see rapid technological advancements, with a focus on improving the stability and reliability of robots in various operational environments [18][19].

Market Overview - Major indices showed positive performance with ChiNext Index up by 2.98%, STAR 50 by 2.51%, Shenzhen Component Index by 2.17%, Shanghai Composite Index by 1.41%, and North Exchange 50 by 1.36% [1] Domestic News - Saiwu Technology announced the completion of the acquisition of Jinlan Nano, aiming to integrate its material R&D and formulation advantages with Jinlan's brand influence and channel network to enhance the market promotion of high-value automotive functional films [1] - A Chinese team successfully developed a high-speed thin-film lithium niobate electro-optic modulator covering near-infrared to mid-infrared bands, achieving signal transmission rates exceeding 200 Gbps and 170 Gbps in O-U and 2-micron bands, respectively [1] - Chipmaker Xinjiun released a new generation of integrated multi-mode communication chip IM3610, achieving full-mode integration of 4G, 5G terrestrial networks, and low-orbit satellite communication, supporting the 3GPP Release 19 standards [1] Overseas News - Hanmi Semiconductor plans to launch a wide-type TC bonding machine for HBM5 and HBM6 production in the second half of this year, with a goal to mass-produce 16-layer and above HBM by 2029 [2] - Intel and AMD have notified Chinese customers about CPU supply shortages, warning that delivery times for some products may extend up to six months [2] - Memory prices have surged by 80%-90% compared to Q4 2025, with DRAM, NAND, and HBM reaching historical highs, and DRAM profit margins expected to exceed HBM [2] - Murata reported a significant increase in orders for capacitors used in AI servers, with total orders reaching 500.7 billion yen, up 11.5% year-on-year, and capacitor orders alone surged by 29.4% to 268.1 billion yen [2] AI Insights - Xiaohongshu is developing an AI video editing product called OpenStoryline, which utilizes natural language commands for automated editing, currently in the testing phase [3] - SenseNova-SI-1.3, an open-source spatial intelligence model by SenseTime, shows significant improvements in core tasks and surpasses previous versions in performance evaluations [3] - AI agents have taken over 80% of enterprise database creation tasks, indicating a transformative impact of AI in infrastructure management [3] - Waymo has introduced a world model based on DeepMind Genie 3, capable of generating realistic and interactive 3D environments for autonomous driving simulations [3] Industry Tracking - Dawn Aerospace's Aurora has been upgraded to a rocket-powered suborbital spaceplane, showcasing rapid testing capabilities and innovation with plans for 15 test flights between 2023 and 2025 [3] - Jingu Co., Ltd. has made significant breakthroughs in its embodied intelligence division, securing multiple mass production orders for robot structural components developed from proprietary materials [3] - Zhirun Medical has developed a flexible electrode that addresses key challenges in brain-machine interface technology, enhancing long-term stability for invasive applications [3] - Owens Corning has launched the SUSTAINA LOOP series of 100% recycled glass fibers, designed to improve the circular economy in the composite materials industry [3]

Core Viewpoint - Pony.ai's stock opened over 7% higher and is currently up 7.33%, trading at HKD 111.30 with a transaction volume of HKD 20.7 million [1] Group 1: Strategic Partnership - Pony.ai and Moore Threads (688795.SH) announced a strategic partnership focused on the implementation and scaling of L4 autonomous driving technology [1] - The collaboration will leverage Pony.ai's core technology, including world models and virtual driver systems, to enhance training and optimization [1] - This partnership represents a significant example of collaborative innovation within China's AI industry, marking the first large-scale application of domestic AI computing power in critical training and simulation processes [1] Group 2: Technological Development - The partnership will utilize Moore Threads' MTT S5000 training and inference integrated computing card and the Quasar computing cluster to advance the training adaptation and validation of Pony.ai's world model and vehicle models [1] - The joint effort aims to build an autonomous driving computing power ecosystem, achieving full-link collaboration across "algorithm-data-computing power-application" [1] - This initiative is expected to accelerate the maturity and cost optimization of L4 autonomous driving technology, providing robust support for the high-quality development of smart mobility and logistics industries [1]