Market Performance - The Hong Kong stock market showed mixed performance with the Hang Seng Index up by 0.06% closing at 26,765.52 points, while the Tech Index fell by 1.24% to 5,725.99 points, and the National Enterprises Index decreased by 0.15% to 9,147.21 points [1][2]. Sector Performance - Traditional cyclical stocks such as oil, coal, and non-ferrous metals performed strongly, indicating a shift in market focus towards value assets. In contrast, previously popular sectors like commercial aerospace and brain-computer interface technologies experienced a pullback [2][11]. Oil Sector - Oil stocks saw significant gains, with China National Offshore Oil Corporation (CNOOC) rising by 4.01%, China Petroleum & Chemical Corporation (Sinopec) increasing by 3.68%, and China Oilfield Services Limited up by 3.23%. This strength is attributed to a recent extreme cold wave in North America, which caused a 63% spike in U.S. natural gas prices from January 20 to 22, subsequently boosting prices of light hydrocarbons and supporting domestic petrochemical prices [4][5]. Coal Sector - Coal stocks also showed robust performance, with Mongolian Energy up by 5.80%, China Shenhua Energy rising by 4.42%, and China Qinfa up by 3.74%. The increase in coal prices is linked to heightened demand due to cold weather, with daily coal output from the Bohai Rim ports averaging 1.682 million tons, a week-on-week increase of 14.2 tons or 9.19% [6][7]. Non-Ferrous Metals Sector - Non-ferrous metal stocks were active, with Rusal rising by 12.31%, China Nonferrous Mining increasing by 9.87%, and Chifeng Jilong Gold Mining up by 7.44%. Citigroup's report indicates a structural bull market in aluminum, predicting a 61,000-ton structural deficit by early 2026 and raising the price target for aluminum from $2,950 to $3,400 per ton [8][9]. Gold Market - Morgan Stanley's research indicates that gold prices have surpassed their forecast of $4,750 per ounce for the second half of the year, with expectations of further increases due to geopolitical risks, central bank purchases, and ETF inflows. A bullish scenario suggests a target price of $5,700 per ounce for the second half [10]. Weakness in Tech Sectors - Commercial aerospace stocks faced collective declines, with Asia Pacific Satellite down by 11.53%, Goldwind Technology down by 7.14%, and JunDa Co. down by 6.56%. This adjustment is attributed to short-term profit-taking pressures [11]. Brain-computer interface stocks also showed weakness, with Nanjing Panda Electronics down by 6.15% and Brainstorming Technology down by 5.35%, partly due to a recent share placement announcement [13][14]. Individual Stock Movements - Zhizhuo (02513.HK) surged by 13.02% following a significant increase in subscription service users after the launch of its flagship model GLM-4.7, indicating strong market interest in its AI capabilities [15]. Lao Pu Gold (06181.HK) rose by 7.80% due to increased customer traffic during a promotional event, with expectations of substantial revenue growth during the upcoming Spring Festival [16].

当前大模型竞争已从"百模大战"进入头部对决阶段，本轮融资及管理团队升级，将为阶跃星辰在技术研 发与市场拓展中提供关键加速度。 据了解，阶跃星辰本轮融资参与方包括上国投先导基金、国寿股权、浦东创投、徐汇资本、无锡梁溪基 金、厦门国贸、华勤技术等产业投资人，腾讯、启明、五源等老股东持续跟投。本轮融资资金将主要用 于基础大模型研发，打造全球顶尖基座模型，并加速"AI+终端"战略的商业化落地。 （编辑：张靖超 审核：李正豪 校对：颜京宁） 中经记者 李静 北京报道 值得关注的是，阶跃星辰同日正式任命印奇为公司董事长，负责战略节奏与技术方向制定。公司核心战 略聚焦"基础大模型"与"AI+终端"，已在语言模型、多模态模型及端云协同领域达到世界领先水平。 1月25日，上海大模型创业公司阶跃星辰阶跃星辰（StepFun）宣布完成超50亿元人民币B+轮融资，刷 新过去12个月中国大模型赛道单笔最高融资纪录。 在业务层面，阶跃星辰已经与OPPO、荣耀、中兴等手机品牌深度合作，模型装机量超4200万台；在汽 车领域，与千里科技、吉利合作推出智能座舱方案，量产车型吉利银河M9上市3个月销量近4万辆。 ...

Core Viewpoint - The article highlights the significant financing achievement of Jumpshare Star, which has raised over 5 billion RMB in a B+ round, surpassing the IPO fundraising targets of competitors like Zhipu AI and MiniMax, indicating strong market confidence in its unique strategy and team [2][3][4]. Group 1: Financing and Capital - Jumpshare Star's recent financing sets a record for single financing in the Chinese large model sector, reflecting a shift in investor focus towards technology barriers, clear commercialization paths, and resilient teams [3][4]. - The diverse shareholder composition, including state-owned enterprises, industrial capital, long-term insurance capital, and market-oriented VC, signals strong confidence in Jumpshare Star's long-term value and stability [4]. - Many investors view Jumpshare Star as their primary or sole investment in the large model field, indicating a concentrated bet on its potential [4]. Group 2: Team and Leadership - The article emphasizes the importance of team composition in Jumpshare Star's success, highlighting a well-rounded team capable of addressing the multifaceted challenges of the large model industry [5][6]. - The appointment of Yin Qi as chairman brings extensive experience in commercializing AI technologies, positioning Jumpshare Star for strategic success [7][8]. - Key team members include Jiang Daxin, who has a strong background in transforming technology into competitive products; Zhang Xiangyu, a leading algorithm scientist; and Zhu Yibo, an expert in AI infrastructure, all contributing to a comprehensive operational strategy [8][9]. Group 3: Strategic Direction - Jumpshare Star's choice of the "AI + terminal" strategy is seen as a unique path amidst the challenges faced by other large model companies, aiming to integrate AI into physical devices [10][11]. - The company focuses on developing a multi-modal AI capability and deep integration with terminal devices, which is essential for creating a comprehensive intelligent assistant [11][12]. - The revenue model is designed to be sustainable, with income linked to the sales of partner manufacturers, indicating a clear and viable business strategy [12][13]. Group 4: Market Performance and Future Outlook - Jumpshare Star has demonstrated strong market demand, with a 170% increase in API call volume over three consecutive quarters and partnerships with major smartphone brands, indicating robust commercial traction [13]. - The company is set to equip over one million new vehicles with its large model technology, showcasing its expansion into the automotive sector [13]. - The article concludes that Jumpshare Star's strategic choices, technological foundations, and business validations position it well for future growth, suggesting a promising trajectory in the AI landscape [14][15].

1月20日，国务院新闻办公室在北京举行新闻发布会，国家发展改革委谈2026年经济工作。国家发 展改革委国民经济综合司司长周陈表示，展望2026年，消费与投资、科技与产业、城乡与区域都将释放 出巨大的发展潜能。从点上看，新技术新产品新场景蔚然成势。我国新能源、新材料、航空航天等新的 经济增长点正在蓄势待发；从线上看，创新链产业链人才链加速融合。我国有完整的产业体系和超大规 模市场，也有全球规模最宏大、门类最齐全的人才资源；从面上看，动力源地区创新引领作用持续增 强。创新集群发展，北京（京津冀）、上海（长三角）、粤港澳大湾区国际科技创新中心。全球科技创 新高地和新兴产业重要策源地正加快打造。 ◆抓住三大潜能实现三个转变 《中国汽车报》：三大潜能对 2026 经济发展将起到怎样的促进作用？又将怎样影响汽车产业发展？ 干春晖：2026年，消费与投资、科技与产业、城乡与区域这三大发展潜能协同发力，将为中国经济 注入持续动力，也给汽车产业带来从做大到做强的关键机遇。说到底，就是通过需求升级带动供给创 新、技术突破改变产业面貌、区域协调放大整体优势，让汽车这个既关系国计民生又代表制造业水平的 行业，在畅通国内循环、参与全球 ...

（原标题：手握最强AI大脑，千里科技打响智能汽车"含模量"战役） 1月26日，大模型领军企业阶跃星辰正式官宣，由印奇出任公司董事长 。这一任命不仅标志着这位跨越 AI双时代的领军人物正式完成其"AI+终端"的战略拼图，更预示着中国版"AI+终端"航母已然起航，加 速推动大模型能力向物理世界的深度渗透 。 据了解，阶跃星辰已顺利完成近50亿人民币的B+轮融资，资金将倾力投入全球领先的基座模型研发， 旨在通过终端Agent的落地，探索AI与硬件融合的新形态。印奇此番同时掌舵阶跃星辰与千里科技，不 仅是两家企业的深度战略接轨，更意味着一个打通底层算法到规模化应用的全链路"Physical AI"体系正 式成型 。 从传感器研究到Physical AI的战略合围 印奇的创业叙事，是一场关于"Physical AI"的长期主义长征。从2011年创办旷视科技起，他便坚持AI的 终极价值在于对物理世界的重构。在早期算法为王的时代，他逆流而上，主张算法必须依附于硬件载体 方能转化为商业价值。从清华姚班到哥大博士，这种"让AI作用于现实"的思维火种始终闪耀。 生态进化，以"含模量"构建差异化技术护城河 "千里+阶跃"战略闭环的 ...

Group 1 - The core viewpoint highlights the acceleration of AI integration within the leading tech companies in the Hong Kong stock market, focusing on the commercialization and application of AI models in various sectors such as e-commerce and transportation [1] - Qianwen APP is advancing productization in key scenarios, while Yuanbao APP's new Spring Red Packet activity is expected to boost downloads and daily active users, indicating a competitive landscape for AI entry points among major model vendors [1] - Open-source securities note that domestic large model vendors are enhancing user and revenue scale through frequent model upgrades and overseas expansion, leveraging social effects from "Spring Festival" activities to capture C-end AI entry points, which may accelerate user growth and commercialization [1] Group 2 - Despite recent fluctuations, the Hong Kong tech sector is seeing active investment in high liquidity products, with the Hang Seng Tech ETF (513130) experiencing a net inflow of 3.609 billion yuan over 13 trading days since January 7, 2026, boosting its scale to 47.586 billion yuan [2] - The Hang Seng Tech ETF (513130) has a daily average trading volume of 5.139 billion yuan, making it the only product tracking the Hang Seng Tech Index with an average daily turnover exceeding 5 billion yuan, showcasing its liquidity advantage [2] - The ETF closely tracks the Hang Seng Tech Index, which includes key competitive tech companies such as SMIC, Alibaba-W, Meituan-W, BYD, and Tencent, all of which are leaders in AI application and development, poised to benefit from the current AI application opportunities [2] Group 3 - The fund manager of the Hang Seng Tech ETF (513130), Huatai-PB Fund, is among the first ETF managers in China, with a strong track record in broad-based and dividend-themed indices [3] - The "Dividend Family" series includes various ETFs that may serve as good defensive options within a "barbell strategy," indicating a diversified approach to investment [3]

1944年，阿根廷作家豪尔赫·路易斯·博尔赫斯曾在《巴别图书馆》中，为人类构筑了一个由无限六角形回廊组成的"全书图书馆"。 起初，它是神谕的代名词。人们沉浸在"万物皆已写就"的狂喜中，穷极一生去追寻那本能解释宇宙真相、甚至是个体命运的"辩护状"。 然而，当狂欢落幕，真相显露：在近乎无限的、由乱码组成的废纸堆里，寻找一行有意义的叙述，概率微弱到近乎虚无。 信息的极度冗余，非但没有点亮智慧，反而以噪声淹没了意义。 博尔赫斯笔下的旅行者穿梭于周而复始的卷册间，看透了那令人绝望的循环。他预言人类终将灭绝，而图书馆永存——"青灯孤照，无限不动，藏有珍 本，默默无闻。" 八十余年后的今天，这个隐喻跨越时空，精准地击中了生成式AI时代的神经。 我们正站在知识饱和的奇点：当算法可以在瞬间穷尽文字的所有组合，我们是在挖掘通往真理的沃土，还是在加速步入一片精神的荒原？当一切皆可生 成，是否意味着一切皆无意义？人类智慧的护城河在何方？我们应该如何与AI共存，并利用它创造出真正有价值的内容？那些记录着人类"不完美灵 魂"的文学与艺术，是否会成为我们最后的精神庇护所？ 图片由AI生成 带着这些关于创造、教育与存在的追问，在高山书院十周 ...

Core Insights - The report "2026 China Technology Mobility Industry 10 Strategic Technology Trends Outlook" aims to provide a key action guide for decision-makers in the industry chain, focusing on cost reduction, enhanced user experience, and ecological collaborative innovation [1][1] - The Chinese technology mobility industry is undergoing a profound systemic transformation driven by the accelerated integration of intelligence, electrification, and artificial intelligence technologies [1][1] - The industry is transitioning from a phase of incremental expansion to a high-quality development stage centered on efficiency, scale, and systemic capabilities amid intensified global automotive competition and evolving profit models [1][1] Trend Summaries - **Trend 1: Chiplet Restructuring Vehicle Chip Architecture** The traditional single-chip SoC is inadequate for L3+ autonomous driving and advanced cockpit requirements. Chiplet architecture is emerging as a mainstream solution for high-performance intelligent driving chips, enabling cost optimization and demand-driven scalability [4] - **Trend 2: AI Box Decoupling Computing Power Deployment** AI Box is becoming a key transitional solution for OEMs to achieve flexible computing power expansion, supporting local AI capabilities without major changes to vehicle electronic architectures [6] - **Trend 3: Accelerated Localization of Automotive Chips** Driven by supply chain security and cost optimization, communication chips and power semiconductors are leading the way in domestic chip localization, transitioning from technical feasibility to large-scale implementation [10] - **Trend 4: Optical Communication in Vehicles** Optical communication is moving from technology validation to mass production, addressing bandwidth limitations and electromagnetic interference in traditional vehicle networks, with significant advancements expected by 2026-2027 [13] - **Trend 5: 48V Low-Voltage Architecture** The 48V architecture is becoming essential for high-power intelligent components in electric vehicles, with initial applications in high-end models expected to scale by 2026 [15] - **Trend 6: Nearing the Inflection Point for Steer-by-Wire** With the maturation of technology and regulatory frameworks, steer-by-wire systems are transitioning from small-scale validation to mass production, becoming critical for advanced driving capabilities [19] - **Trend 7: Large Models Driving Intelligent Cockpits** The evolution of large models is transforming intelligent cockpits into system-level intelligent entities, enhancing user interaction and service optimization [22] - **Trend 8: Phased Implementation of L3 Autonomous Driving** L3 autonomous driving will be introduced gradually under regulatory constraints, focusing on safety and responsibility alignment, with commercial applications expected to begin in 2024-2025 [25] - **Trend 9: Redefining Interaction with Small Screens** The shift from centralized large screens to distributed small screens in vehicle interfaces aims to enhance user experience and reduce cognitive load during driving [28] - **Trend 10: Physical AI Driving Second Growth Curve** Physical AI capabilities are evolving into transferable competencies, enabling automotive technologies to extend across various intelligent terminals, fostering ecosystem collaboration [30] Overall Industry Outlook - The trends indicate a systemic restructuring in the Chinese technology mobility industry, focusing on integrated systems engineering and collaborative efficiency, with a shift from isolated performance to comprehensive system capabilities [32][34] - The next phase of competition will favor organizations that can build replicable, scalable, and evolvable system capabilities, marking 2026 as a pivotal year for technological scaling and global competitive landscape reshaping [34]

2026年1月26日，行业领先的通用大模型创业公司阶跃星辰（StepFun）正式宣布完成超50亿元人民币B+轮融资，一举创下过去12个月中国大模型领域单 笔最高融资纪录。同日，人工智能领域领军人物印奇出任公司董事长，与现有核心管理团队携手，全面加速"基础大模型+AI+终端"双轮战略落地，为国 产大模型商业化竞争注入强音。 本次B+轮融资阵容堪称豪华，汇聚了"国资引导+产业资本+头部创投"的多元力量。上国投先导基金、国寿股权、浦东创投、徐汇资本等国资及产业投资 人强势入局，厦门国贸、华勤技术等产业方深度参与，腾讯、启明、五源等老股东则持续加码跟投，彰显了资本市场对阶跃星辰技术实力与商业化路径的 高度认可。 据介绍，本轮超50亿元融资将重点投入基础模型研发，聚焦全球顶尖基座模型的技术攻坚，同时加速"AI+终端"战略的场景落地与生态拓展。在大模型行 业从"百模大战"迈入"头部对决"的关键阶段，这笔创纪录的融资不仅为阶跃星辰提供了充足的资金储备，更构建起覆盖资本、产业、技术的协同生态，为 其在核心技术研发与商业化竞争中奠定优势。 与巨额融资同步落地的，是公司核心管理团队的重磅升级。人工智能领域知名创业者印奇正式出任阶 ...

Core Insights - The large model sector in China is experiencing significant investment, with Jieyue Xingchen completing over 5 billion yuan in Series B+ financing, setting a record for single financing in the past 12 months [1] - The funding will be used for foundational model research and to accelerate the implementation of the AI + terminal strategy [1] - Jin Qi, chairman of Qianli Technology, has been appointed as chairman of Jieyue Xingchen, focusing on strategic direction and technological development [2] Investment and Financing - Jieyue Xingchen's recent financing round included participation from various institutional investors such as Shangguotou Xiandao Fund and Tencent, among others [1] - The company aims to develop a top-tier foundational model and enhance its AI + terminal strategy through this funding [1] Leadership and Strategic Direction - Jin Qi has extensive experience in AI and automotive integration, having previously led AI applications in various real-world scenarios [2][4] - His appointment is seen as a strategic move to strengthen the collaboration between Jieyue Xingchen and Qianli Technology, focusing on AI applications in automotive technology [2][4] Technological Development - Jieyue Xingchen has released three generations of foundational large models, with the latest model achieving high inference efficiency and focusing on multimodal capabilities [6][7] - The company is also targeting deployment across multiple platforms, including mobile, automotive, and IoT devices, with significant partnerships already established [7][9] Market Impact and Future Plans - Jieyue Xingchen's models have been integrated into over 42 million devices, serving nearly 20 million users daily [9] - The company anticipates that its large models will be deployed in over one million vehicles this year, indicating strong market penetration [9] - Jin Qi emphasizes the importance of integrating AI with terminals in three core scenarios: individual, transportation, and home, with a focus on vehicles as the primary entry point [9]