市场喧嚣期待之时，知名科技评论作者Kevin Xu发布长文对DeepSeek的商业模式、护城河进行了分析， 在AI圈内引起了很大反响。 在即将到来的1月27日——也就是"DeepSeek时刻"一周年之际，全球AI社区都期待DeepSeek再来个大 招。 他认为，DeepSeek最坚固的护城河，是它零外部融资、无商业化压力的独特模式。在全球AI巨头都被 资本裹挟着必须赚钱的时候，DeepSeek是唯一一个可以不计成本、不看脸色、只为AGI梦想狂奔的"自 由人"。 Kevin S. Xu（徐凯文）是专注中美科技与资本交叉领域的独立观察者，知名科技评论人。他创办的 ChinaTalk播客及Newsletter在专业圈层颇具影响力，擅长从资本流向、组织行为与地缘政治多维度解构 技术演进逻辑。 我们给大家梳理了一下文章要点： 给市场预期泼盆冷水 ：虽然大家都在等DeepSeek的新模型，但作者直言"不要指望它能像去年那样再次震惊世界"。因为现在 的市场已经被"开源模型"喂饱了，DeepSeek虽然打响了第一枪，但现在并不是唯一、也不是最开源的玩 家了（比如它至今没开源数据集）。 唯一的"零融资"异类 ：现在的AI圈就是 ...

市场喧嚣期待之时，知名科技评论作者Kevin Xu发布长文对DeepSeek的商业模式、护城河进行了分析， 在AI圈内引起了很大反响。 他认为，DeepSeek最坚固的护城河，是它零外部融资、无商业化压力的独特模式。在全球AI巨头都被 资本裹挟着必须赚钱的时候，DeepSeek是唯一一个可以不计成本、不看脸色、只为AGI梦想狂奔的"自 由人"。 Kevin S. Xu（徐凯文）是专注中美科技与资本交叉领域的独立观察者，知名科技评论人。他创办的 ChinaTalk播客及Newsletter在专业圈层颇具影响力，擅长从资本流向、组织行为与地缘政治多维度解构 技术演进逻辑。 我们给大家梳理了一下文章要点： 作者原文如下（由AI翻译）： 没有商业模式：DeepSeek的长期优势 作者：Kevin S. Xu 在即将到来的1月27日——也就是"DeepSeek时刻"一周年之际，全球AI社区都期待DeepSeek再来个大 招。 随着1月27日"DeepSeek时刻（DeepSeek Moment）"一周年纪念日的临近，市场对于 DeepSeek在农历新年（2月17日）前发布一款更强大新模型的期待正日益高涨。 然而，过高的 ...

第一篇论文（mHC）出来的时候，打开论文的人都表示很懵，直呼看不懂，让 AI 助手用各种方式讲给自己听。我们也翻了翻网友的讨论，发现理解起来 比较透彻的办法其实还是要回到研究脉络，看看这些年研究者们是怎么接力的。要理解第二篇论文（Conditional Memory）也是如此。 2026 年 1 月过半，我们依然没有等来 DeepSeek V4，但它的模样已经愈发清晰。 最近，DeepSeek 连发了两篇论文，一篇解决信息如何稳定流动，另一篇聚焦知识如何高效检索。 于是，我们就去翻各路研究者的分析。这个时候，我们发现了一个有意思的现象：DeepSeek 和字节 Seed 团队的很多工作其实是存在「接力」的 —— mHC 在字节 Seed 团队 HC（Hyper-Connections）的基础上进行了重大改进；Conditional Memory 则引用了字节 Seed 的 OverEncoding、UltraMem 等 多项工作。 如果把这些工作之间的关系搞清楚，相信我们不仅可以加深对 DeepSeek 论文的理解，还能看清大模型架构创新正在往哪些方向突破。 在这篇文章中，我们结合自己的观察和学界专家的点评， ...

专题：2025微博财经之夜暨北京财经大V联盟年会 2025微博财经之夜暨北京财经大V联盟年会于1月15日在北京举行。前东北证券首席经济学家付鹏出席 并演讲。 付鹏在演讲中提到了2015、2016年的两个历史事件：第一，当年埃隆·马斯克的SpaceX失败，他含泪的 画面，是对人类文明勇于挑战、颠覆的纪念；第二，投资圈大家熟悉的木头姐，2015年她那张经典 PPT，把人类文明未来所有技术路径都列在上面，告诉大家这就是未来，让大家把钱给她，她来投资。 "大家称她为女版巴菲特，我一直跟很多人说她不是，她和巴菲特完全不同，她更像二级市场里的一级 市场投资人。"付鹏说。 他认为，作为产业、技术，早期投资所谓泡沫并非坏事，需要这样冒风险的资本。人类历史往前翻一百 多年，扬帆远航需要资本支持，可能是国王、贵族支持，需要有愿意冒风险、钱、船、人都回不来的人 去挑战、支持。所以，整个产业生命周期必然经历早期一级市场投资、估值泡沫、估值杀泡沫阶段。 付鹏进一步指出，生产关系本身还需做很多事，比如以人为本，对居民部门加大福利、补偿等。只有这 样，才能保证世界秩序不再次大规模崩塌。所以要聚焦于生产力前行，同时也要聚焦改变各国内部生产 ...

Core Viewpoint - DeepSeek, a Chinese startup, is set to launch its new generation model V4 around mid-February 2026, aiming to make a significant impact during the Chinese New Year period [1]. Group 1: Company Development - DeepSeek has shown remarkable growth over the past two years, launching its foundational model V3 on December 26, 2024, and an open-source inference model R1 on January 20, 2025, which gained significant attention for its explicit reasoning capabilities [4]. - The R1+V3 chat product has also received high domestic recognition, establishing DeepSeek as a benchmark enterprise in China's AI engineering capabilities [4]. Group 2: Model V4 Features - The V4 model is designed to significantly enhance programming capabilities, achieving a record score of 92.0 in authoritative programming benchmarks like Design2Code, surpassing products from leading overseas companies such as GPT-4.5 and Claude3.7 [6]. - A key breakthrough of V4 is its ability to handle ultra-long context processing, utilizing an NSA mechanism to achieve a 6-9 times speed increase under a 64K context window, allowing it to process millions of tokens effectively [6]. Group 3: Technical Innovations - V4 was developed under constraints of high-end GPU availability, addressing common issues in large model training such as performance degradation through innovative technical methods rather than relying solely on computational power [7]. - The introduction of the mHC architecture has significantly improved training stability, with a mere 6.7% increase in training time leading to a rise in accuracy for complex reasoning tasks from 43.8% to 51.0% [7]. Group 4: Research Contributions - On January 12, DeepSeek published a new training architecture paper co-authored by its founder and researchers from Peking University, introducing the Engram conditional memory module, which decouples computation from storage [9][10]. - This approach allows for model scaling without relying on an increase in chip quantity, providing a new technical pathway for AI companies constrained by hardware limitations [10]. Group 5: Industry Context - The large model landscape has become increasingly competitive, with open-source becoming a core trend in 2025, as both large enterprises and startups strive for dominance in the global open-source ecosystem [11]. - The launch of V4 transcends mere product iteration, serving as a "technical examination" to validate DeepSeek's technological leadership and the maturity of its architectural innovations [13]. Group 6: Market Implications - The performance of V4 will not only impact DeepSeek's standing in the global open-source ecosystem but also reflect the maturity of China's large model technology route [16]. - The ongoing competition has shifted from a focus on parameter counts to the intricacies of technical methods and operational efficiency, indicating a new phase in the industry [16].

Core Insights - DeepSeek, a Chinese AI startup, is set to launch its next-generation AI model V4 in mid-February, which is expected to outperform competitors like Anthropic's Claude and OpenAI's GPT series [1] - The rapid development of AI models in China, particularly by DeepSeek, has significantly narrowed the gap with the US in the AI sector over the past year [2] Group 1: Company Developments - DeepSeek's R1 model was launched last year and completed training in just two months at a fraction of the cost incurred by US companies, achieving comparable performance to ChatGPT and Meta's Llama [2] - Chinese open-source AI models account for nearly 30% of global AI technology usage, with companies like Alibaba's Qwen model gaining traction among developers worldwide [3] - Alibaba has released nearly 400 open-source models, with over 18 million downloads, showcasing its significant role in the global AI landscape [3] Group 2: Competitive Landscape - The US AI strategy focuses on high-end capabilities, closed-source models, and platform products, while China's approach emphasizes open-source, engineering efficiency, and rapid industrial deployment [4][5] - While the US leads in cutting-edge model capabilities, China excels in engineering efficiency and speed of implementation, with no significant time lag in these areas [5] Group 3: Future Trends - The next significant advancements in AI are expected to occur in areas such as humanoid robots integrated with large models, industrial AI models for complex processes, and breakthroughs in low-cost inference and edge computing [10] - The AI toy industry is projected to reach a milestone of 1 million units sold, which will generate substantial interaction data, enhancing the AI models' capabilities and establishing AI toys as essential items in daily life [11]

年初至今，一项衡量中国内地科技股的指数已上涨12%，表现跑赢香港以及美国的同类指数，因投 资者纷纷涌入。从芯片到人形机器人再到商用火箭等领域的每日进展，以及大量计划中的股票上市，共 同推动了这股热潮。 展望未来，Rasid认为人工智能支出和更有利的政策将成为推动中国科技股的关键催化剂。 "我们确实认为中国在科技领域仍然有很多机会。"该公司的全球市场策略师Raisah Rasid在新加坡 的一次简报会上表示："你们将会看到越来越多机器人技术的进步，以及更多DeepSeek时刻。" 摩根资产管理表示，随着中国加大力度创建更多类似DeepSeek的公司，中国科技股将继续受益于 技术突破。 ...

Summary of Conference Call on AI in Healthcare - DeepSeek Industry Overview - The application of AI in the healthcare sector is increasingly widespread, particularly in areas such as image recognition, Clinical Decision Support Systems (CDSS), and intelligent triage [1][2] - Collaboration between Ruijin Hospital and Huawei has led to the development of an open-source large model that significantly enhances pathological recognition capabilities, which has been adopted by multiple hospitals to alleviate the shortage of pathologists [1][2] Key Insights and Arguments - AI technology has shown remarkable progress in medical imaging, especially in chest X-rays, CT scans, MRIs, and angiography, with significant efficiency in identifying lung nodules during the pandemic [2] - The integration of AI in rehabilitation robotics is a promising area, particularly in community hospitals, with companies like Fourier Intelligence making substantial advancements [2][3] - The current AI systems in hospitals are primarily provided by external vendors, while hospitals supply the necessary hardware, such as Haiguang CPUs and Huawei 910B integrated machines [5] - AI accounts for approximately 1% of total healthcare IT spending, with one-third of that allocated to AI solutions [6] Emerging Trends - Personal health applications, both domestically (e.g., Ant Financial's Aifuku) and internationally (e.g., OpenAI Health), are rapidly developing, focusing on managing patient data through apps in collaboration with healthcare professionals [7][8] - Future data integration efforts may focus on chronic diseases like diabetes, with third-party platforms facilitating data sharing and utilization [9] Data Management and Integration - Current efforts in hospital data management are being spearheaded by local health authorities, with projects in the planning stages to organize and potentially trade data assets [10][11] - Although no hospital has fully established a comprehensive data management system yet, pilot projects are underway to explore data asset trading [11] Competitive Dynamics - The relationship between public hospitals, third-party companies, and large enterprises is evolving, with commercial entities potentially addressing service limitations imposed by insurance reimbursement standards [12] - The demand for rehabilitation services is high in aging cities like Shanghai, where the shortage of rehabilitation physicians and expensive equipment presents challenges [3][14] Future Prospects - The acceptance of large models in hospitals has increased significantly, with AI technology becoming a standard component in various healthcare IT projects [4] - The integration of AI in hospital management is expected to enhance operational efficiency and improve service quality [4] - The market for rehabilitation robots is expected to diversify in terms of payment models, with potential for private institutions to adopt service fees or insurance payments [17] Conclusion - The healthcare industry is on the brink of a significant transformation driven by AI technologies, with ongoing developments in data management, rehabilitation robotics, and personalized health applications paving the way for improved patient care and operational efficiency [1][2][3][4][5][6][7][8][9][10][11][12][14][17]

Core Insights - DeepSeek, a Chinese AI startup, is set to launch its next-generation AI model V4 in mid-February, which is expected to outperform competitors like Anthropic's Claude and OpenAI's GPT series [1] - The rapid development of AI in China has narrowed the gap with the US, with experts noting that the progress made in just one year is significant [1][2] Group 1: Company Developments - DeepSeek's R1 model was launched last year and completed training in just two months at a fraction of the cost incurred by US companies, achieving comparable performance to ChatGPT and Meta's Llama [2] - Chinese open-source AI models account for nearly 30% of global AI technology usage, with companies like Airbnb and Meta utilizing models developed by Alibaba [3] - Alibaba has released nearly 400 open-source models, with over 18 million derivatives and 700 million downloads, showcasing its significant role in the global AI landscape [3] Group 2: Competitive Landscape - The US AI strategy focuses on high-performance closed-source models and platform products, while China emphasizes open-source models and rapid industrial application [4] - While the US leads in cutting-edge model capabilities, China excels in engineering efficiency and speed of deployment, with no significant time lag in these areas [5] Group 3: Future Trends - The next significant advancements in AI are expected to occur in areas such as humanoid robots integrated with large models, industrial applications, and breakthroughs in low-cost inference and edge computing [10] - The AI toy industry is projected to reach a milestone of 1 million units sold, which will generate substantial interaction data, enhancing model capabilities and establishing AI toys as essential daily items [11]

摩根资产管理表示，随着中国加大力度创建更多类似DeepSeek的公司，中国科技股将继续受益于技术 突破。"我们确实认为中国在科技领域仍然有很多机会。"该公司的全球市场策略师Raisah Rasid在新加 坡的一次简报会上表示："你们将会看到越来越多机器人技术的进步，以及更多DeepSeek时刻。" ...