21世纪经济报道记者冉黎黎北京报道2026年将成为"人工智能+"应用落地元年——这是12月29日举行的 赛迪顾问"2026年IT趋势"发布会上给出的研判。 受访赛迪顾问专家指出，此前人工智能虽已有一些比较好的落地，但"人工智能+"行业应用整体还在探 索阶段，而2026年政府及央国企有望在政务、工业、能源等重点领域加快布局AI应用，率先形成一批 高价值、可复制推广的应用场景示范项目。 从算力来看，赛迪顾问电子信息产业研究中心资深分析师张耀嵘在发布会上介绍，近年来，随着大模型 与行业应用深度融合，算力需求呈现指数级增长。预计2026年中国算力总规模将突破1200 EFLOPS，稳 居全球第二。其中，智能算力作为规模增长的核心引擎，贡献率接近90%。 随着"人工智能+"行动深入推进，算力需求将大幅增长。赛迪顾问预计2026年中国算力总规模将突破 1200 EFLOPS，稳居全球第二。其中，智能算力作为规模增长的核心引擎，贡献率接近90%。与此同 时，相关商业模式也正发生转变，无论是人工智能应用还是算力服务，在2026年，商业模式都有望 从"为技术或功能付费"向"为效果付费"转变。 需要注意的是，2025年8月印发的 ...

为持续引导投资者树立股东意识，积极行使股东权利，提升上市公司投资者关系管理水平，营造理性投资、价值投资、长期投资的市场氛围，9月17日，东 方证券联合上海证券交易所、上海市证券同业公会联合开展"我是股东"——走进沪市上市公司招商轮船（601872）活动顺利举行。东方证券投资者教育基地 组织高净值个人及机构投资者等30余位，实地参观了轮船招商总局，并与公司高管进行了深度交流。 "我是股东"走进上市公司活动自2013年启动以来，已陪伴投资者走过十余年历程，累计组织投资者走进沪市上市公司超过2000家次，仅2024年就达到450家 次，成为资本市场促进投资者与上市公司双向沟通、增强投资者获得感的重要品牌。 在投资者交流会现场，招商轮船董事会秘书孔康，董事会办公室总经理李漫、董事会办公室副总经理刘宇丰、ESG总监蔡晓华等出席活动。招商轮船董秘孔 康先生在致辞中向投资者重点介绍了公司近期在航运主业和ESG工作中取得的亮眼成绩，并强调公司始终将投资者回报和持股体验放在公司价值创造的核心 位置，让广大股东切实分享公司发展的成果。 据了解，招商轮船作为招商局旗下重点发展远洋运输的航运企业，经营和管理着大中华地区历史最悠久、最 ...

Core Viewpoint - Beijing Zhiyu Huazhang Technology Co., Ltd., known as "China's version of OpenAI," is set to launch its IPO on the Hong Kong Stock Exchange, aiming to raise approximately HKD 4.3 billion with an expected market capitalization exceeding HKD 51.1 billion upon listing [1][2]. Group 1: IPO Details - The company plans to issue 37.42 million H-shares, with 5% allocated for Hong Kong and 95% for international investors, plus an overallotment option of 15% [1]. - The expected share price is HKD 116.20 per share, with a minimum purchase of 100 shares [1]. - The IPO subscription period runs from December 30, 2023, to January 5, 2024, with the official listing scheduled for January 8, 2024 [1]. Group 2: Financial Backing and Investment - Zhiyu has completed eight rounds of financing prior to the IPO, raising over HKD 8.3 billion [2]. - The company has secured cornerstone investors, including JSC International Investment Fund SPC and JinYi Capital Multi-Strategy Fund SPC, with a total subscription amount of approximately HKD 2.98 billion [2]. Group 3: Business Overview - Founded in June 2019, Zhiyu focuses on developing general large models and has launched China's first proprietary pre-trained model framework, GLM [4]. - The company has supported over 8,000 institutional clients and approximately 800,000 devices as of June 30, 2025 [4]. - Revenue figures for 2022, 2023, and 2024 are projected at HKD 57 million, HKD 125 million, and HKD 312 million, respectively, reflecting a compound annual growth rate of over 130% [4]. - Despite significant revenue growth, the company remains unprofitable [4]. Group 4: Market Position - According to Frost & Sullivan, Zhiyu ranks first among independent general large model developers in China and second overall, holding a market share of 6.6% based on 2024 revenue [4]. - The competitive landscape includes other large model companies, such as MiniMax, which recently went public [4].

Core Insights - The establishment of the "Medical Artificial Intelligence Innovation Joint Research Center" by Peking University People's Hospital and Ant Group Health aims to advance the application of AI technologies in the healthcare sector [1][2] - The launch of the first national standard for "AI doctors" in the surgical field signifies a systematic step towards standardizing the technical application of medical AI [1][2][3] Group 1: AI Research and Development - The research center will focus on addressing clinical pain points and exploring innovative applications of AI in specialized disease diagnosis, clinical decision support, and health management models [1] - The GAPS (Grounding, Adequacy, Perturbation, Safety) evaluation framework for large models in specialized disease evidence-based capabilities has been developed and applied in the "Antifufu" App [2] Group 2: Collaboration and Standards - Over 500 doctors have contributed to the "Famous Doctor AI Avatar" feature on the "Antifufu" App, providing 24/7 professional health consultation services to the public [2] - The national standard for "AI doctors" in the surgical field outlines technical requirements for medical expertise, interactive service capabilities, safety, and ethical compliance [3] Group 3: Industry Impact - The integration of AI in clinical decision support and health management is becoming increasingly important, necessitating clear capability boundaries, unified technical standards, and an evaluable governance mechanism for scalable industry development [2] - Ant Group emphasizes the importance of collaboration with industry partners to drive forward research and innovative applications of AI in healthcare [3]

大家好，我是佩妮～ 今天是个好日子！撒花 除了 Manus 被 Meta 收购，其实 AI 行业还有个大事件，就是 大模型公司智谱华章，在港交所开始申购 了，预计在1月8日在港交所主板上市。 以每股 116.2 港币发行价计算，这次的募资总额预计达到 43 亿港币，市值超过 511 亿港币 （约 460 亿RMB 左右）。 | 以每股 | 116.2 港币发行价计算，这次的募资总额预计达到 | 43 亿港币，市值超过 | 511 亿港币 | （约 | 460 亿RMB 左右）。 | | | | --- | --- | --- | --- | --- | --- | --- | --- | | （也是好消息，和Manus撞上了有点生不逢时，堪称 | AI 界汪峰……） | | | | | | | | AI 行业的朋友，前阵肯定都会刷到 | MiniMax 这两家明星公司，通过港交所聆讯的消息。 | 只要关注 | 智谱 | 和 | 前阵子还老有人猜谁会是第一股（智谱是 | 19 号披露的招股书，minimax 晚了两天），现在不用猜了，花落智谱，不愧是 | "AI 国家队" 。 | | IPO 都喜欢叫自己 | ...

Core Viewpoint - The company, Zhihua Technology, is launching its IPO with an expected total fundraising of HKD 4.3 billion, aiming to be the first global stock focused on general AI models, with a market capitalization projected to exceed HKD 51.1 billion [1][3]. Group 1: IPO Details - Zhihua plans to issue 37.42 million H-shares, with 1.87 million shares for public offering in Hong Kong and 35.55 million shares for international offering, priced at HKD 116.20 per share [1]. - The IPO is set to begin on December 30, 2025, and is expected to conclude on January 5, 2026, with the listing on the Hong Kong Stock Exchange scheduled for January 8, 2026 [1]. Group 2: Investment and Market Position - Zhihua has completed eight rounds of financing totaling over HKD 8.3 billion, attracting notable investment institutions as cornerstone investors, who plan to subscribe for HKD 2.98 billion, accounting for nearly 70% of the offering [3]. - The company ranks first among independent general model developers in China and second among all general model developers based on projected revenue for 2024 [3]. Group 3: Fund Utilization - Approximately 70% of the net proceeds from the IPO (around HKD 2.9 billion) will be allocated to research and development of general AI models to strengthen the company's competitive edge [4]. - About 10% (approximately HKD 420 million) will be used to optimize the company's MaaS platform, enhancing its foundational models and infrastructure [4]. Group 4: Revenue Growth and Business Model - Zhihua's revenue primarily comes from its MaaS (Model as a Service) model, which has shown exponential growth since its launch in 2021, with a significant increase in revenue contribution from this segment [5][6]. - The company reported revenues of HKD 57.4 million, HKD 124.5 million, and HKD 312.4 million for 2022, 2023, and 2024 respectively, achieving a compound annual growth rate of 130% [6]. Group 5: R&D Investment and Technological Leadership - Zhihua has consistently increased its R&D investment, with expenditures of HKD 84.4 million, HKD 528.9 million, and HKD 2.1954 billion for 2022, 2023, and 2024 respectively, totaling around HKD 4.4 billion [7]. - The company has developed the GLM series of models, which are updated every 3-6 months, maintaining a leading position globally [7]. Group 6: Market Outlook - The global AI market is projected to exceed USD 500 billion by 2027, with generative AI and related services expected to account for over 30% of this growth [8]. - Zhihua's listing marks a significant milestone for the AI model industry in China, entering a new phase driven by both technology and capital [8].

人民财讯12月30日电，据京东云消息，近日，京东与华为技术有限公司正式达成深度技术合作。双方将 聚焦智能广告、大模型及云计算等核心领域，通过京东云计算、广告算法升级、华为鲲鹏技术联合创 新，重点破解高并发场景下的广告技术瓶颈，共同构建下一代高效、智能的广告技术基础设施。 ...

Core Insights - The acquisition of Manus by Meta is not a game-changer or a critical move; it is part of a broader trend of mergers and acquisitions in the AI industry expected to intensify by 2026 [1][7] - The central theme of competition is the "force" behind AI models, which is essential for creating intelligent applications [1][5] Group 1: Acquisition Analysis - The purchase of Manus is not a decisive factor for Meta, similar to the previous $14.8 billion acquisition of Scale AI [8] - Manus may help Meta in user engagement but lacks the "force" that drives innovation; the acquisition focuses on the team's ability to leverage models rather than acquiring core technology [3][9] - Meta's AI strategy is currently weak in foundational models, which are crucial for its competitive positioning [2][10] Group 2: Future Directions - Meta is expected to pursue significant acquisitions in the range of $20 billion to $50 billion to strengthen its model capabilities [3][11] - The company is also likely to recruit entire teams or individual talents from leading AI firms to enhance its technological prowess [3][11] - The AI landscape is anticipated to undergo major transformations by 2026, with significant mergers and acquisitions centered around the concept of "force" [12][13] Group 3: Competitive Landscape - The AI industry is entering a phase of intense competition, with major players engaging in strategic mergers, talent acquisition, and ecosystem collaborations [6][12] - The competition will not only reshape the AI model landscape but also impact traditional industries and employment structures [12][13] - Companies lacking the "force" in their AI capabilities, including major tech firms, are likely to seek acquisitions to bolster their positions [5][13]

Core Insights - The report by CCID Consulting outlines the development blueprint and transformation directions for the IT industry in China by 2026, focusing on ten key areas including computing power, storage, AI, industrial software, and financial digital transformation [1][2][3]. Computing Power - The demand for computing power is expected to grow exponentially due to the deep integration of large models and industry applications, with China's total computing power projected to exceed 1200 EFLOPS by 2026, maintaining the second position globally, and intelligent computing contributing nearly 90% [1]. - A collaborative model of "cloud centralized training and edge flexible reasoning" will be widely adopted, with super-node architectures overcoming traditional computing cluster bottlenecks, becoming a significant choice for large intelligent computing centers [1]. Artificial Intelligence - The "AI+" sector is anticipated to enter a new phase of typical demonstration and large-scale implementation by 2026, with cloud services accelerating towards AI-native evolution, transitioning from "resource supply" to "intelligent engine" [2]. - High-quality private data sets will become scarce resources, and lightweight AI terminals will serve as crucial carriers for application implementation, driving deep penetration of "AI+" in industries such as automotive and logistics [2]. - The large model market is expected to experience explosive growth, with both deployment and usage volumes increasing significantly, focusing on vertical small models and autonomous collaborative intelligent agents [2]. Industrial Software - The global industrial software market is projected to grow steadily from 2023 to 2025, with a compound annual growth rate of approximately 6.4%, reflecting a balance of stable growth and structural optimization [2]. - By 2026, industrial software is expected to play a vital role in transforming manufacturing from experience-driven to data-driven, enhancing industry efficiency and resource allocation [2]. Financial Digital Transformation - By 2025, the financial industry is expected to undergo a high-quality paradigm shift from "digitalization" to "intelligentization," driven by AI and open API ecosystems, supported by distributed core system restructuring and technological breakthroughs [3]. - IT investment in China's financial sector is projected to continue expanding, with AI reshaping financial technology and generative AI accelerating penetration into vertical financial fields, evolving IT from a supportive tool to a core engine driving financial innovation [3].

文 | 大力财经 快手的技术核心圈，年末传来重磅人事变动！ 据多位知情人士爆料，快手科技副总裁、基础大模型及推荐模型负责人周国睿已进入休假状态，内部系统签名已改为"Log Out"，这意味着这位执掌快手核 心推荐算法的技术大牛，即将正式告别快手。消息一出，立刻引发行业震动——毕竟周国睿不是普通高管，他亲手主导了快手推荐大模型的革新，是程一笑 力推AI战略的关键执行者。 要搞懂这场人事变动的分量，得先看清周国睿在快手的"含金量"——他手里攥着的，是快手的"生命线"。 而这波离职潮的背后，是快手进入"程一笑时代"后的组织架构大调整。2023年底，快手宣布了上市以来最大规模的组织架构和管理层调整，涉及主站、电商 与商业化三个事业部的15条业务线，25名中层管理人员受到影响。架构调整必然伴随业务合并和权责变动，不少高管因此被架空，这也被认为是近期高管密 集离职的核心原因。 除此之外，快手内部治理的漏洞也被暴露过。今年9月初，曾有举报邮件流出，称快手某员工泄露公司核心数据和规划打法获利，给公司造成重大损失。这 一事件也让外界对快手的内部管控产生质疑，或许也加剧了核心人才的流失。 2021年加入快手后，周国睿很快坐上核心 ...