生成式AI 一、Claude Cowork新增11款插件，AI直接 威胁 SaaS业务流 1. Anthropic为Claude Cowork新增11款插件，覆盖销售、财务、法律等领域，AI直接取代SaaS软件端到端完成业 务工作流； 2. 全球软件股遭遇"SaaS末日"抛售，一周内蒸发近万亿美元市值，Gartner暴跌21%、Thomson Reuters跌18%、 ServiceNow跌11%； 3. 行业正从按席位收费的SaaS模式向按产出计费的AaaS（Agent即服务）模式转变，传统软件护城河面临底层模型 公司的降维打击。 https://mp.weixin.qq.com/s/UoYyAlxPkdkcoPnNrvNP5g 二、GitHub集成Claude和Codex，与Copilot形成三足鼎立 1. GitHub正式集成Claude和Codex，与Copilot形成AI编程"三足鼎立"，开发者可通过Agent HQ一站式指挥三个AI 协同工作； 2. 开发者可在同一个编码难题上同时指派三个AI异步执行，对比不同方案，支持IDE、GitHub网页端和移动端一键调 用； 3. GitHub从代码托管平 ...

"当全球创作者开始使用我们的技术标准、遵循我们的生态规则时，中华文化的国际影响力将实现系统 性跃升。"刘伟表示， 此外，随着出海进入"深水区"，企业面临的挑战已从单纯的市场竞争，深化为知识产权保护、全球合规 与黑灰产治理等复杂的系统性挑战。刘伟表示，上海作为链接全球的桥头堡，在营商环境法治化方面走 在前列。 近年来，上海多次开展针对数字侵权、海外私服及盗版黑色产业链的专项治理，有效维护本土企业的全 球合法权益。建议进一步先行先试，构建对接国际高标准经贸规则的"全链条维权机制"：搭建一站式全 球知识产权保护服务平台，建立海外维权预警体系，为企业在国际竞争中提供坚实的法治保障，通过优 化营商环境"软实力"，构筑起抵御全球风浪的"避风港"和"加速器"。 回顾上海游戏产业的"出海群像"，从米哈游到叠纸、鹰角、沐瞳等企业，"上海出品"之所以能赢得全球 市场，共性在于都选择"技术立身"的发展道路：以"高精尖制造"级的标准打磨产品，以工业化管线保障 高质量内容的持续产出，海外用户往往先被高精度的实时渲染、物理仿真带来的"技术震撼"所吸引，进 而才会沉浸其中，理解其中蕴含的东方哲学与中华文化。 他建议，上海在政策层面应当像支 ...

格隆汇2月5日｜据澎湃，上海市人大代表、米哈游创始人、总裁刘伟表示，上海游戏产业已具备世界一 流的工业化管线能力，不应止步于做全球市场的"内容供应商"，而应致力于成为全球数字文化生态 的"架构师"。刘伟认为，要利用上海在人才、资本和产业链上的集聚优势，主动参与全球数字文化规则 的制定。从数字资产的确权标准，到元宇宙空间的交互协议，再到生成式AI的伦理规范，上海企业应 在政府引导下，勇于在国际舞台发出"上海声音"。 ...

证券代码：002131 证券简称：利欧股份 利欧集团股份有限公司投资者关系活动记录表 编号：2026-004 | | ■特定对象调研 □分析师会议 | | --- | --- | | 投资者关系活 | □媒体采访 □业绩说明会 | | | □新闻发布会 □路演活动 | | 动类别 | □现场参观 □一对一沟通 | | | □其他 | | | 华鑫证券 朱珠、何春玉、朱丹宁、黄英豪；浙江丰道投资 夏 | | | 甘霖；上海杭贵投资 饶欣莹；西安敦成私募基金 张晓艳；长 | | | 城财富资产 乔笙；泽铭投资 单河；福泽源 (上海) 私募基金 | | 参与单位名称 | 林秋；上海磐松私募基金 张潇；上海睿源私募基金 田华；第 | | 及人员姓名 | 一北京投资 梁卓均；上海颢升私募基金 夏敬慧；苏州君榕资 | | | 产管理 熊晓峰；深圳熙山资本管理 张亮；厦门中略投资管理 | | | 肖晶晶；深圳市前海夷吾资产管理 谢树平；共青城鼎睿资产 | | | 管理 王可丹；上海名禹资产管理 刘宝军；爱建证券 章孝林 | | 时间 | 年 月 日 2026 2 5 | | 地点 | 电话会议 | | 上市公司接待 | 董 ...

Core Viewpoint - Huang Renxun, CEO of Nvidia, expressed that programming is merely typing and has lost its value, emphasizing the need for innovation and adaptation in the face of rapid technological evolution [4][6][20]. Group 1: AI and Innovation - Huang advocates for a management approach that encourages experimentation and exploration within organizations, stating that innovation often occurs outside of strict control [6][7][40]. - He emphasizes the importance of allowing teams to experiment with various AI tools without the immediate pressure of ROI calculations, fostering a culture of "letting a thousand flowers bloom" [6][40]. - The concept of "AI factories" is introduced, where the focus shifts from merely creating tools to generating digital labor, fundamentally transforming industries [9][10][68]. Group 2: Market Potential - Huang highlighted the vast market potential for AI, noting that the global IT industry is approximately $1 trillion, while the total global economy is around $100 trillion, indicating a hundredfold opportunity for AI to penetrate and reshape the remaining sectors [10][11][68]. - He pointed out that every industry has the chance to transform into a technology-driven company, with examples like Disney aspiring to be like Netflix and Mercedes wanting to emulate Tesla [11][68]. Group 3: New Paradigms in Computing - Huang discussed the shift from explicit programming to implicit programming, where users can simply express their intentions and the AI will generate the necessary code, thus lowering the technical barrier [20][70]. - He described the transition from a "pre-recorded" era of software to a "generative" era, where applications will be contextually unique and dynamically generated based on user interactions [17][52]. Group 4: Data Sovereignty and Knowledge Ownership - Huang stressed the importance of data sovereignty and the need for companies to maintain control over their core intellectual property, which he believes lies in the questions posed during interactions with AI rather than the answers themselves [16][73]. - He argued that the most valuable intellectual property is the ability to ask the right questions, which reflects a company's strategic thinking and innovation potential [16][73]. Group 5: Future of Work and AI Integration - Huang predicted that future companies will have numerous AI assistants that learn from employee decisions and inquiries, ultimately becoming unique intellectual assets for the organization [16][73]. - He warned that companies must adapt quickly to AI technologies, as those who fail to do so risk being outpaced by competitors who embrace these advancements [68][73].

Core Viewpoint - The article discusses the strategic transformation of Chinese retail brands going global, driven by "integrated commerce" and "intelligent commerce" in response to globalization and digitalization trends [5][23]. Market Trends - Consumer demand is evolving towards a more personalized, efficient, and diverse shopping experience, necessitating a seamless shopping experience centered around consumers [7][24]. - Integrated commerce, which combines multiple online and offline channels into a unified service loop, is becoming a core competitive advantage for retail brands [7][24]. - Despite many companies having established multi-channel layouts, data silos hinder effective channel collaboration and operational closure, posing a significant barrier to the implementation of integrated commerce [7][24]. Technological Changes - The rapid development of generative AI is catalyzing the widespread adoption of "intelligent commerce," which reshapes consumer processes and expands sales scenarios while significantly enhancing operational efficiency [9][25]. - Intelligent commerce is deeply integrated with integrated commerce, allowing brands to deliver highly efficient and personalized seamless experiences, creating a "unified × intelligent" effect that opens new growth opportunities [9][25]. - Currently, many companies are still in the exploratory phase of applying AI, with untapped potential in marketing recommendations, supply chain optimization, and payment risk control [9][25]. Core Foundation - Achieving integrated and intelligent commerce relies on the comprehensive integration of data across all dimensions, with payment data serving as a crucial link between transactions, users, and scenarios [12][27]. - Establishing a "unified and intelligent payment" system allows brands to upgrade payment processes from mere transaction endpoints to data hubs and strategic nodes, laying the groundwork for business integration and intelligent decision-making [12][27]. Payment Challenges and Solutions - Companies face three main challenges in payment: optimizing payment experience, managing security and compliance, and reducing operational costs while increasing efficiency [16][28]. - Leveraging professional third-party payment services is seen as the most cost-effective solution for growing companies, enhancing payment success rates and customer satisfaction through integrated global payment methods and AI-driven personalized recommendations [16][28]. - Intelligent risk control and anti-fraud systems based on vast transaction data can help brands effectively mitigate data breaches and cross-border compliance risks [16][28]. Key Findings from Research - 32% of companies identified data security and compliance risks as their top challenge in cross-border payments, followed by inconsistent payment experiences across channels (21%) and the need for multiple currency support (18%) [18][60]. - 90% of companies believe that data sharing between user and order data across channels will enhance repurchase rates, with 40% expecting significant improvements [43][44]. - 67% of companies believe that a consistent payment experience is key to improving sales conversion rates, while 87% think that one-click payment features will significantly boost repurchase rates [63][64]. Future Directions - To adapt to changing consumer behaviors in the global market, retail brands must create seamless shopping experiences and achieve integrated commerce [83]. - Companies should focus on consumer-centric channel restructuring, breaking down payment and order data silos to maximize the value of intelligent commerce [83]. - Utilizing professional third-party payment services can optimize payment experiences, reduce compliance risks, and drive overseas business growth [84].

市场表现 -50 0 50 100 150 200 250 300 (%) 协创数据 沪深300 资料来源：Wind，华鑫证券研究 2026 年 02 月 05 日 公司业绩表现亮眼，AI 算力硬件与 AI 算力服务 业务双轮驱动 买入(首次) 事件 分析师：何鹏程 S1050525070002 hepc@cfsc.com.cn 联系人：石俊烨 S1050125060011 shijy@cfsc.com.cn | 基本数据 | 2026-02-04 | | --- | --- | | 当前股价（元） | 205.85 | | 总市值（亿元） | 712 | | 总股本（百万股） | 346 | | 流通股本（百万股） | 345 | | 52 周价格范围（元） | 71.28-231.68 | | 日均成交额（百万元） | 1719.5 | 相关研究 —协创数据（300857.SZ）公司事件点评报告 协创数据发布业绩公告：预计 2025 年实现归母净利润 10.5 亿 元至 12.5 亿元，同比增长 51.78%至 80.69%%；实现扣非净利 润 10.1 亿元至 12.1 亿元，同比增长 50.45%至 ...

随着 De epSe ek等大模型技术 的持续爆发 ， 生成式AI和大模型技术呈现爆发式增长， 算 法工程师 岗位再度迎来 " 黄金爆发期 " 。 确保你学习完以后，能够具备在实际工作当中 解决实际问题的能力，成为一名合格的AI算法工程师。 并且该计划承诺在校应届生就业薪资拿不到29W、 在职人员最低薪资涨幅低于40%-50% ，均可以全额退款 。 目前已经有上千位同学通过该计划拿到offer了， 平均薪资达到35w+， 最新一期学员就业薪资最高能拿到 85 K ！ 扫码添加顾问，备注 "咨询课程" 2026届校招数据显示，大模型算法工程师 月薪中位数 已逼近 3万元 ，顶尖人才的年薪 破百万 ，远超 前后端、 运维、嵌入式 等传统 技术岗位，成为程序员群体中的"薪资天花板"。 AI的快速发展也给很多程序员带来 更多的机会 ， 很多公司都 投入了巨量资源， 把 人工智能 作为下一阶段的主要发展方向， 字节，腾讯，京东等 大厂增加了很多 AI部门 ，扩招大量的 大模型和算法人才 。 所以在 大模型、生成式AI 这些领域，各家公司都给足了诚意，像 DeepSeek最近热招的 大模型全栈工程师 等 AI核心岗位 更 ...

Core Insights - Google DeepMind's launch of Project Genie has raised concerns about the impact of AI tools on the gaming industry, leading to significant stock declines for companies like Unity and Take-Two [1][3] - Project Genie is a generative AI model capable of creating real-time rendered, interactive 3D virtual scenes, but it still faces limitations in editing capabilities and detail refinement [1][3][4] - The gaming industry is recognizing the revolutionary potential of AI models, but there are ongoing debates about their application and the associated compliance issues [2][6] Group 1: Market Reactions - Unity's stock fell by 24.22%, marking its largest single-day drop since 2022, while Take-Two and Roblox also experienced significant declines of 7.93% and 13.17%, respectively [1] - The release of Project Genie has sparked intense discussions within the industry, with many companies expressing mixed feelings about AI's role in game development [6][7] Group 2: AI Model Limitations - Project Genie currently lacks "editable" attributes, making it difficult for developers to adjust scenes and interactions, which are crucial for the competitive edge of open-world games [3][4] - The model's early-stage development means it does not yet significantly impact the business models or competitive landscape of game engines and developers [4] Group 3: Potential Benefits - Smaller and independent game developers may leverage Project Genie to enhance their product quality and reduce the gap with larger firms [5] - The real-time generation capability of AI could expand the possibilities for game formats, particularly in open-world exploration games, enhancing user engagement [5][12] Group 4: Industry Perspectives - A significant portion of game developers (over 50%) believe that generative AI has a negative impact on the industry, a sentiment that has increased compared to two years ago [7] - Some companies, like Games Workshop and Hooded Horse, have implemented strict policies against the use of AI in content creation, reflecting a cautious approach to AI integration [7][8] Group 5: Future Directions - The industry is moving towards a more nuanced understanding of AI's role, shifting from a binary "yes or no" stance to more refined operational models [10] - There is a call for the development of AI application platforms and support systems to facilitate the integration of AI into game development processes [11]

Group 1 - Nvidia is nearing a $20 billion investment agreement to participate in OpenAI's latest funding round, marking Nvidia's largest single investment to date, with CEO Jensen Huang stating "this is a very good investment" [1] - OpenAI's current funding round aims for a total of $100 billion, with Amazon planning to invest up to $50 billion and SoftBank considering a $30 billion investment, leading to an estimated valuation of approximately $830 billion [1] - This investment signifies a deeper integration between AI infrastructure and leading model developers, with capital increasingly concentrating among a few super players [1] Group 2 - Tencent has officially open-sourced its high-performance LLM inference core operator library, HPC-Ops, built from scratch using CUDA and CuTe, achieving a 30% improvement in inference QPM for the Mix Yuan model and a 17% improvement for the DeepSeek model [2] - In terms of single-operator performance, Attention shows up to a 2.22x improvement over FlashInfer/FlashAttention, while GroupGEMM outperforms DeepGEMM by up to 1.88x, and FusedMoE exceeds TensorRT-LLM by up to 1.49x [2] - The operator library is optimized for mainstream inference graphics cards in China, addressing high usage costs and hardware compatibility issues with existing mainstream operator libraries [2] Group 3 - Alibaba has open-sourced the Qwen3-Coder-Next model, featuring 80 billion parameters with only 3 billion active parameters, achieving over 70% problem-solving rate on the SWE-Bench Verified, comparable to models with 10-20 times more active parameters [3] - The model excels in long-sequence reasoning, complex tool usage, and recovery from execution failures, supporting a context of 256k and seamless integration with various IDE platforms like Cline and Claude Code [3] - A paper co-authored by Zhou Jingren and Lin Junyang has been published alongside the SWE-Universe framework, expanding the real-world multilingual SWE environment to nearly one million levels [3] Group 4 - The website rentahuman.ai has launched, allowing AI to hire humans for tasks such as delivery, event check-in, and on-site inspections through the MCP protocol or REST API [4] - Within 48 hours of launch, the platform had over 20,000 available human workers, allowing individuals to set their own hourly rates without the need for small talk, with tasks including photography, restaurant tasting, and package collection [4] - The site has sparked discussions on responsibility attribution, task authenticity verification, and the ethics of AI hiring humans, also seen as a demonstration of the MCP protocol's value [4] Group 5 - Mianbi Intelligence has open-sourced the MiniCPM-o 4.5 model, which features only 9 billion parameters and achieves full-duplex dialogue capabilities, becoming the first large model for "instant free conversation" [5] - The model employs an end-to-end multimodal architecture, utilizing time-division multiplexing and active interaction mechanisms to automatically decide whether to speak at a frequency of 1Hz, ensuring continuous perception and dynamic dialogue [5] Group 6 - Kunlun Tiangong has released the Skywork desktop version, which executes tasks locally without uploading to the cloud, capable of reading vast local files for summarization and new product generation while supporting parallel multitasking [6] - It supports switching between Claude Opus 4.5, Sonnet 4.5, and Gemini 3 Pro models, with over 100 selected skills built-in, covering Office suite, web pages, and image and video generation [6] - The application prioritizes Windows systems, offering higher quality image and video generation, with all operations conducted in a local virtual machine environment to ensure data security [6] Group 7 - Apple has released Xcode version 26.3, officially introducing "intelligent agent programming" support, allowing developers to directly call AI agents like Anthropic's Claude and OpenAI's Codex [7] - The integrated AI agents can browse and search the entire project structure, read, write, edit, and delete files, and automatically reference Apple's official documentation to resolve issues [7] - User feedback has been mixed, with some praising the experience while others report issues such as freezing, poor diff mechanisms, and instability in cross-file refactoring [7] Group 8 - The open-source music generation model ACE-Step 1.5 has gained support on ComfyUI, utilizing a hybrid LM+DiT architecture to generate a complete 4-minute song in approximately 1 second on an RTX 5090 [8] - The model supports over 50 language instructions and can run with less than 4GB of VRAM, achieving a music coherence score of 4.72, surpassing most commercial models [8] - It allows for LoRA fine-tuning for style personalization and will soon support music reconstruction and segment repair features, all running locally to ensure data security [8] Group 9 - Google has launched PaperBanana, establishing a multi-agent collaborative framework for generating paper illustrations, aimed at freeing researchers from time-consuming illustration tasks [9] - The system includes roles such as retriever, planner, modeler, visualization expert, and critic, achieving improvements in simplicity, readability, and overall aesthetic quality [9] - However, there are limitations in handling complex architectures, such as text distortion or connection errors, with plans to introduce code diffusion models for drawing and human-machine collaboration interfaces in the future [9]