深圳商报首席评论员 胡蓉 企业之间的良性竞争，是用户的幸事。期待AI应用的竞争不会急功近利，同时要防止"AI幻觉"误导用 户。只要能为用户带来真正的价值，市场也将予以丰厚的回报。 2023年被称为"中国大模型元年"，大量企业机构涌入赛道。经过两年的发展，据新华社在2025世界人工 智能大会上了解到的信息，截至今年7月，我国已发布1509个大模型，在全球已发布的3755个大模型中 数量居于首位。这个数字不可谓不惊人。其发布主体则包括互联网巨头、AI公司、高校与科研机构和 传统龙头企业等。 人工智能技术将对产业变革、经济发展以及社会运行带来巨大改变，这已成共识，因此如此之多的企业 和机构急切涌入AI大模型赛道不难理解。但随着时间的推移，投入不菲的AI大模型如何赚钱盈利的问 题，也变得越来越迫切。 与当年互联网应用程序走过的发展道路一样，AI应用同样将从跑马圈地走向市场搏杀；与互联网技术 能够突破时空界限一样，AI应用同样不会出现"山头林立、群雄共存"的局面。因为用户的关注度是有限 的，正如当下在即时通信、网络购物等领域所展现的一样，AI应用在各个领域也都会走向"行业巨头"把 持。 另一方面，公众已经历了对AI大模 ...

Core Insights - The competition for the title of "the first global large model stock" has officially begun, with two Chinese AI unicorns, MiniMax and Zhiyu AI, reportedly passing the Hong Kong Stock Exchange listing hearing this week [1] - Both companies have received approval from the China Securities Regulatory Commission for overseas listing and participated in the Hong Kong hearing [1] - If all goes well, these companies will become the fastest cases to pass the hearing since the new filing system for mainland companies to list in Hong Kong was implemented [1] Company Performance - Zhiyu AI, established in 2019, has achieved significant progress in business and financing, with an annual recurring revenue (ARR) exceeding 100 million RMB [2] - The company serves over 2.7 million paying customers through its API platform, including several large domestic tech firms [2] - Zhiyu AI's new AI-driven coding tool subscription plan has gained over 150,000 users, priced at 20 RMB per month, significantly lower than competitors [2] - The company anticipates over 100% revenue growth by 2025 [2] Financing and Investment - MiniMax, founded in 2021, completed a $600 million Series A funding round in March 2024, with a post-money valuation of $2.5 billion, led by Alibaba [3] - The investor lineup for MiniMax includes prominent firms such as Alibaba, Sequoia, and Tencent [3] - MiniMax's new text model, MiniMax-M2, ranks in the top five globally on the Artificial Analysis (AA) leaderboard, marking a significant achievement for Chinese open-source models [3]

Core Insights - Xiaomi has officially launched its self-developed AI model Xiaomi MiMo-V2-Flash, marking a significant step towards the "Agent era" in AI technology [1] - The company reported a revenue growth of 32.5% year-on-year, exceeding 340 billion yuan, with an adjusted profit increase of 73.5% for the first three quarters of the year [1] - Xiaomi plans to invest over 200 billion yuan in R&D over the next five years, with an estimated 40 billion yuan allocated for 2026 [1] Company Developments - The Xiaomi ecosystem includes personal devices, transportation devices, and home devices, supported by core technologies such as chips, operating systems, and AI [2] - The new head of Xiaomi's MiMo model, Luo Fuli, has publicly confirmed her role at Xiaomi, previously being a key member of DeepSeek [2] - Luo Fuli emphasized the vision of creating a true "intelligent agent" that understands and coexists with the world, rather than just a language simulator [2] Product and Market Strategy - Xiaomi's MiMo series includes various models such as MiMo-7B, MiMo-VL, MiMo-Audio, MiMo-VL-Miloco, and MiMo-Embodied, showcasing a diverse range of AI capabilities [1] - The company has established a global ecosystem with over 1 billion active devices and coverage in more than 100 countries [3] - Xiaomi's internet services have developed a comprehensive ecosystem, with significant advancements in short drama content, including 1,000 S-level short dramas [3]

巨头的入局往往是一条赛道成熟的标志，而Rokid、雷鸟创新、影目科技等创业公司是这条赛道的铺路者，如今他们相逢于"AI眼镜元年"。 近日，2025微博年度关键词出炉，是一个古早又新鲜的词汇：赛博。 这个词诞生已近百年，在互联网语境中更是历经多轮传播与演变，其核心含义早已无需过多解释，还衍生出了层出不穷的网络新梗；说它新鲜，则源于其 承载的独特内核——永远锚定着对未来天马行空的想象。 曾经承载"赛博"之名的科幻畅想，一旦从虚幻照进现实，便会褪去"赛博"的标签，而更前沿、更具未来感的构想，又会迅速接过"赛博"的话语权，让这个 词始终保持着对未知的探索与浪漫。 产品功能设计与产业链生产能力的成熟，让智能眼镜迅速进入销量增长的通道。IDC报告显示，2025年上半年全球智能眼镜市场出货量达406.5万台，同比 增长64.2%；2025年全球智能眼镜出货量将达到1205万台，同比增长18.3%；其中中国市场将突破275万台，同比增长107%，居全球第一。 2025年11月至今，联想、百度、阿里、理想接连推出了AI眼镜，加上此前就已入局的华为、小米，以及曾从腾讯手中抢下PICO的字节跳动，AI眼镜赛道 上已经聚集了互联网大 ...

Core Viewpoint - The article discusses Tencent's significant upgrade in its AI model development framework, highlighted by the appointment of renowned AI scholar Vincesyao as Chief AI Scientist, indicating a strategic shift towards systematic engineering in AI model development [2][5]. Group 1: Key Personnel and Strategic Shift - Vincesyao, a former OpenAI scientist, joins Tencent to lead AI Infra and the large language model department, reporting directly to Tencent's president [2][5]. - His expertise in AI agents and large model reasoning is expected to enhance Tencent's capabilities in AI, aligning with the company's focus on systematic engineering and AI infrastructure [5][6]. Group 2: Structural Upgrades - Tencent has established three key departments: AI Infra, AI Data, and a data computing platform, to strengthen its large model development foundation [6][8]. - The AI Infra department will focus on distributed training and high-performance inference services, while the AI Data department will concentrate on data and evaluation systems [6][8]. Group 3: Competitive Landscape - The article emphasizes that AI competition is evolving beyond model parameters to a "system war" that integrates data, infrastructure, and algorithms [8]. - Tencent's internal AI efficiency transformation has led to the deployment of its mixed Yuan model in over 900 applications and scenarios [10]. Group 4: Achievements and Performance Metrics - Tencent's mixed Yuan model has released over 30 new models in the past year, with the latest mixed Yuan 2.0 leading in complex reasoning and text generation [13]. - The AI capabilities have been integrated into major products like WeChat and QQ, with 90% of Tencent engineers using the CodeBuddy AI code assistant, generating 50% of new code with AI assistance [13].

Group 1 - Tencent has upgraded its large model research and development architecture by establishing new departments: AI Infra, AI Data, and Data Computing Platform, to enhance its core capabilities in large model development [2] - Vincesyao, a former researcher at OpenAI, has been appointed as the Chief AI Scientist, overseeing the AI Infra and Large Language Model departments, reporting directly to Tencent's President [2] - The AI Infra department will focus on building technical capabilities for large model training and inference platforms, emphasizing distributed training and high-performance inference services [2] Group 2 - The AI Data and Data Computing Platform departments will be responsible for constructing the data and evaluation systems for large models, as well as integrating big data and machine learning into a data intelligence platform [2] - Over the past year, Tencent has released more than 30 new models under its Mix Yuan large model initiative, with over 900 applications and scenarios implemented internally across various platforms such as Tencent Meeting, WeChat, and games [3] - More than 90% of Tencent engineers are using the Tencent Cloud Code Assistant, CodeBuddy, with 50% of new code being generated with AI assistance [3]

此次调整也是腾讯在今年紧锣密鼓的AI布局中，颇为重磅的一步。 36氪独家获悉，腾讯近期完成了一次组织调整，正式新成立AI Infra部、AI Data部、数据计算平台部。 12月17日下午发布的内部公告中，腾讯表示，Vinces Yao将出任"CEO/总裁办公室"首席 AI 科学家，向 腾讯总裁刘炽平汇报；他同时兼任AI Infra部、大语言模型部负责人，向技术工程事业群总裁卢山汇 报。 腾讯并未披露Vinces Yao的中文名或过往履历。不过，36氪了解到，Vinces Yao即为数月前入职腾讯的 姚顺雨，他毕业于清华和普林斯顿大学，曾任OpenAI研究员，是OpenAI首批智能体产品Operator与 Deep Research的核心贡献者。 此前，混元大模型团队虽是腾讯的公司级项目，拉通了各个BG的不同板块，就TEG内部而言，参与到 混元模型研发的就有大预语言模型部、AI Lab、机器学习平台等等部门。经过调整后，模型团队内部力 量会更加统一。 新成立的 AI Data 部、数据计算平台部，将分别负责大模型数据及评测体系建设、大数据和机器学习的 数据智能融合平台建设工作。 其中，王迪继续担任大语言模型部 ...

Core Insights - The article highlights the introduction of Xiaomi's new AI model, Xiaomi MiMo-V2-Flash, by AI expert Luo Fuli, who recently joined Xiaomi from DeepSeek [2][4] - Luo Fuli emphasizes that AI is evolving in a non-linear manner, paralleling the evolution of the human brain over 600 million years [2][15] - Xiaomi MiMo-V2-Flash, with a total parameter count of 309 billion, ranks as the second-best open-source model globally, outperforming competitors in speed and cost efficiency [5][10] Company Overview - Xiaomi's AI model, Xiaomi MiMo-V2-Flash, was launched at the "Human-Vehicle-Home Ecosystem" partner conference, showcasing the company's commitment to AI development [2][4] - The model's core team is described as a young, innovative group that integrates research, product development, and engineering [2][4] Model Specifications - Xiaomi MiMo-V2-Flash features 309 billion total parameters and 15 billion active parameters, specifically designed for intelligent agents [5][10] - The model achieves a threefold increase in inference speed compared to DeepSeek-V3.2 while being more cost-effective [11][12] Technical Innovations - The development of Xiaomi MiMo-V2-Flash focused on three key issues: efficient communication capabilities, high inference efficiency, and the transition from pre-training to post-training paradigms [7][8][9] - The model incorporates innovative structures such as a mixed Sliding Window Attention and Global Attention, along with Multi-Token Prediction training to enhance performance [12][16] Future Vision - Luo Fuli articulates a vision for the next generation of intelligent systems that must coexist with the physical world, emphasizing the need for memory, reasoning, and decision-making capabilities [15][16] - The goal is to create a virtual universe with physical consistency and temporal coherence, representing a significant leap in AI capabilities [16]

Group 1 - The core viewpoint indicates a significant increase in domestic and international excavator sales in November 2025, despite a relatively weak real estate demand in China [1] - Domestic fixed asset investment in infrastructure continues to show a downward trend, while mining fixed asset investment has seen a slight recovery over the past two months [1] - The average working hours and operating rates of major engineering machinery products have shown a month-on-month improvement in November [1] Group 2 - The penetration rate of engineering machinery in the mining sector is expected to increase, driven by the commencement of major national projects, accelerated funding, and replacement policies [1] - High demand persists in Africa and regions along the Belt and Road, while demand in Europe and North America is gradually recovering [1] - The U.S. government is focusing on the development of humanoid robots, intending to accelerate policy issuance to boost industry growth [1] Group 3 - Domestic central and local governments continue to issue policies to promote the development of humanoid robots, alongside advancements in AI models and dexterous hands, which are expected to accelerate the industrialization process [1] - The Robot Industry ETF (159551) tracks the Robot Index (H30590), which focuses on the robotics and automation sectors [1] - The index selects listed companies involved in industrial and service robots and their supply chains, reflecting the overall performance of the robotics industry [1]

Core Insights - The article discusses the potential transformation of mobile phones into accessories for AR glasses, highlighting how future interactions may be revolutionized [2] - It features an interview with AR industry pioneer Zhu Mingming, who explains the fundamental differences between VR and AR, and how AI large models can empower smart glasses [2] - Zhu believes that China has already gained a leading position in this sector and encourages young people to leverage their youth as a powerful asset, warning that stagnation poses the greatest risk [2] Industry Overview - The conversation emphasizes the significance of augmented reality (AR) and virtual reality (VR) technologies, particularly in the context of smart glasses [2] - The integration of digital information with reality through AI advancements is a key focus, suggesting a shift in how users will interact with technology [2] Market Position - Zhu Mingming asserts that China is at the forefront of the AR industry, indicating a competitive advantage in the global market [2] - The encouragement for young individuals to innovate and adapt reflects a broader trend of valuing agility and creativity in technology development [2]