Core Insights - The recent upgrade of DeepSeek does not include multimodal visual understanding capabilities, focusing instead on pure text and voice interaction paths [2] - The core context window has been increased from 128K to 1M tokens, allowing the model to process long texts equivalent to the "Three-Body Problem" trilogy in a single instance, positioning it against international competitors like GPT-5 and Gemini3Pro [2] - The knowledge base of the current model has been updated to include accurate outputs for news events as far ahead as April 2025, with the cutoff date for knowledge now set to May 2025 [2] User Experience and Development - Feedback from developers and early users indicates that the new model's language style has become "enthusiastic and nuanced," with front-end response quality rated as comparable to Claude3.5Sonnet, suggesting a focus on enhancing user interaction experience [5] - The company has been actively hiring for multiple technical core positions, including deep learning researchers and engineers, indicating a commitment to advancing its large language model (LLM) capabilities [5] - There is speculation within the industry that the current version may correspond to the rumored "DeepSeek V4" or an enhanced version of the V3.2 series, although the official version name has not yet been disclosed [5]

凤凰网科技讯2月11日，继被曝出全新V4系列架构规划后，国产大模型头部玩家深度求索今日傍晚逐步 商量了一个关键更新。 除参数层级的硬核跃升外，本次灰度测试还释放出显著的体验层信号。多名开发者及早期体验者反馈， 新模型的语言风格明显趋向"热情而细腻"，前端响应质感被评价为"可媲美Claude3.5Sonnet"。这或是 DeepSeek在保持极致推理性能的同时，对C端交互体验发起的一次精准补课。 值得注意的是，此番升级并未包含多模态视觉理解能力，模型仍聚焦于纯文本与语音交互路径。 凤凰网科技此前曾发展，DeepSeek在近期曾大量招人。放出了多个技术核心岗位招聘，包括深度学习 研究员(负责LLM各个方向的前沿探索，如预训练、alignment、codemath、backbone、multimodal等)、 深度学习研发工程师(LLM预训练数据/搜索策略相关)、大模型全栈工程师(算法与工程并重，归属于各 个前沿探索方向，除了实现还会参与一些环境的部署和搭建)、全栈开发工程师、核心系统研发工程 署。且据招聘者表述，校招、社招、实习均开放，另有产品、设计、数据百晓生等岗位开放。 截至目前，深度求索官方尚未就此轮灰度发 ...

Core Insights - Huawei Cloud officially launched "CodeArts," an AI-powered coding assistant, in January 2023, which integrates IDE, autonomous development mode, and code library indexing capabilities, currently in public beta for 10,000 users [1][8] - The personal version of "CodeArts" is available for free to developers, while the enterprise version will be announced later [1][8] - The product utilizes GLM-4.7 and DeepSeek-V3.2 models and supports JetBrains series and Visual Studio Code IDEs [1][8] Product Features - "CodeArts" combines essential programming capabilities such as project-level code generation, code continuation, research knowledge Q&A, and unit test case generation, significantly enhancing developer productivity and providing a high-quality coding experience [2][9] - The tool allows users to input requirements, enabling the AI to generate code directly [3][11] Copyright and Usage - Huawei Cloud states that the copyright of the code generated by the AI belongs to the user, emphasizing that "CodeArts" functions as a tool that responds to user inputs without creative autonomy [5][13]

接下来，影视行业的一切价值都将重估。 摘要： 一年前的春节，一家中国公司用R1大模型震撼全球科技界。凭借算法创新突破算力限制，以较低训练成本达到以往AI大模型靠堆算力、拼资金和数据的 效果，外媒将这一突破称为"DeepSeek时刻"。 那场技术地震的余波尚未平息，2026年2月，字节跳动似乎正准备接棒，在视频生成领域制造另一场海啸。 就在几天前，字节跳动旗下的最新一代视频生成模型Seedance 2.0悄然在即梦、豆包等产品中开启小范围内测。 如果你还记得OpenAI发布Sora时的那个夜晚，那么现在的感觉似曾相识，却又有所不同。不同之处在于，这一次，站在风暴眼中心的是又一家中国公 司，更重要的是，这不再是少数人手中的演示Demo，而是一个正在被数万创作者疯狂测试的"实战武器"。 一位曾专注大银幕创作的从业者向凤凰网科技感叹，"Seedance 2.0的完成度令人惊艳，尝试了猫狗宠物的创作，效果都很好"，但其也对影视行业接下来 的发展感到担忧，"能预见制作层面的预算将被砍，《流浪地球3》这样的作品几年前立项成本数亿，这种还在制作过程中的作品，接下来所有的一切都要 价值重估"。 一枚王炸，悄悄扔出 2月7日晚间 ...

Core Insights - The article discusses the imminent release of new AI models in the Chinese market, particularly focusing on the GLM-5 model from Zhipu AI, which is expected to leverage advanced technologies and compete effectively in the AI landscape [1][16]. Group 1: Model Development and Features - The GLM-5 model has been linked to multiple technical platforms, indicating a strong collaborative effort in its development [2][4]. - GLM-5 incorporates a 78-layer Transformer decoder with a total parameter count of approximately 745 billion, which includes a mixture of dense and sparse architectures [6][8]. - The model utilizes a hybrid expert (MoE) architecture, activating only a small fraction of its parameters during inference, which enhances efficiency while maintaining performance [9][10]. Group 2: Technological Innovations - The integration of DeepSeek's Sparse Attention (DSA) mechanism allows GLM-5 to handle long sequences more efficiently, reducing computational costs significantly [12][13]. - Multi-Token Prediction (MTP) technology is employed to accelerate token generation, allowing the model to predict multiple tokens simultaneously, which is particularly beneficial for structured text generation tasks [15][16]. - The model's architecture reflects a shift towards efficiency over sheer parameter count, indicating a trend in the AI industry towards optimizing performance rather than simply increasing size [9][17]. Group 3: Market Position and Challenges - GLM-5 is expected to excel in code generation and logical reasoning tasks, positioning it competitively in software development and algorithm design [16]. - However, the model currently lacks multi-modal capabilities, which may limit its applicability in creative AI-generated content (AIGC) scenarios, especially as competitors advance in this area [16]. - The article highlights a broader industry trend where companies are moving towards open-source technology integration, emphasizing efficiency and practicality in AI model development [16][17].

微软高管曾公开表示，抓取互联网内容训练AI属于"合理使用"。这种数据殖民主义思维，揭示了技术巨 头对数据主权的漠视。企业投入大量资源积累的数据资产，可能正成为他人模型的养料。 赛凡企业云盘为企业建立完全自主的AI运行环境。所有数据从生成、存储到处理全程留在企业内部， 彻底杜绝敏感信息外泄风险。 硬件一体化设计让部署变得简单。根据模型规模需求，赛凡提供从单卡RTX 4090到多卡A100/H100的完 整解决方案。企业无需担心复杂的硬件兼容性问题，真正实现开箱即用。 DeepSeek在企业内部的深度集成，展现出强大潜力。一家金融机构部署后，风控模型准确率提升11%， 而所有数据始终在合规边界内流动。 物理层面，数据存储在企业自有机房，从根源确保控制权。相比公有云无法知晓数据具体位置的情况， 私有云提供完全透明性。 当AI开始阅读你的合同、分析你的客户数据、处理你的财务报告时，你是否想过这些敏感信息正在"离 家出走"？ AI时代的数据主权危机正在悄然蔓延。许多企业为提升效率而拥抱AI，却无意中将核心数据暴露于第 三方平台。员工将销售合同粘贴至公共AI界面，财务数据通过API传输至云端，研发图纸被上传至在线 模型— ...

生意社02月09日讯 中国铝业股份有限公司2026年2月9日铝锭(AL99.70)现货价格各地区价格大幅上涨，具体如下： 华东市场对外报价23400元/吨，华南市场对外报价23430元/吨，西南市场对外报价23370元/吨，中原市 场对外报价23280元/吨；较上一交易日(2.6)分别上涨260元/吨、290元/吨、260元/吨、250元/吨。 PriceSeek评析 生意社02月09日讯 中国铝业股份有限公司2026年2月9日铝锭(AL99.70)现货价格各地区价格大幅上涨，具体如下： 华东市场对外报价23400元/吨，华南市场对外报价23430元/吨，西南市场对外报价23370元/吨，中原市 场对外报价23280元/吨；较上一交易日(2.6)分别上涨260元/吨、290元/吨、260元/吨、250元/吨。 PriceSeek评析 铝，多空评分：1 文章显示2026年2月9日铝锭(AL99.70)现货价格在华东、华南、西南、中原市场分别上涨260元/吨、290 元/吨、260元/吨、250元/吨，涨幅显著。这表明市场对铝的需求强劲或供应紧张，利好现货价格。铝作 为大宗商品，价格上升可能受下游制造业复苏或库存 ...

南方财经全媒体记者 林汉垚 实习生 徐若萱 国家金融监督管理总局最新披露的数据显示，2025年保险业原保险保费收入首次突破6万亿元大关。与此同时，与之配套的 数字化服务正在经历一场由生成式AI引领的供给侧改革。 中国保险行业协会此前发布的《中国保险业社会责任报告（2024）》显示，保险业正加快数字化转型，2024年AI坐席服务 量已达9.37亿次。行业数智化进程正在从"效率工具"向"决策辅助"跨越。保险消费者对复杂保单的解构需求日益增长，利用 生成式人工智能（AIGC）进行保单分析、核保咨询及方案规划逐渐成为新趋势。 近期，21世纪经济报道记者以普通消费者身份，针对百万医疗险条款、家庭保障设计及复杂健康告知等真实场景，对 DeepSeek、腾讯元宝、通义千问、Kimi、豆包等国产主流大模型进行了实测。 测试结果显示，大模型在"条款解读"方面表现卓越，能将长达万字的保险合同精准提炼为易读的免责清单，极大地降低了 消费者的阅读门槛。 但在专业深度层面，大模型分析仍存偏差。北京大学应用经济学博士后朱俊生教授指出，AI目前更适合作为前端知识工具 和辅助决策支持系统，而非独立的保险咨询或销售主体。 可降低消费者认知门槛 ...

Group 1 - The recent promotion by Qianwen App allowing users to purchase milk tea for 0.01 yuan signifies a pivotal moment for AI applications in China, moving from simple chat functions to actionable tasks [1] - OpenClaw, an open-source AI agent, has gained rapid popularity among developers, indicating a shift in AI applications towards more practical uses [2] - Investment analysts are optimistic about AI as a key market trend, with expectations for significant technological breakthroughs in AI multimodal applications and biomedical fields over the next few years [2][3] Group 2 - The emergence of "AI-native super applications" like LiblibAI is anticipated to create benchmark cases of user growth and revenue by 2026 [3] - The AI application market is transitioning from infrastructure investment to commercial value realization, with a focus on cash-generating applications in the downstream sector [3] - Hardware innovations in smartphones, cars, and smart home devices are expected to drive sustainable industry trends, with companies like Apple leading the charge [3] Group 3 - The field of embodied intelligence, particularly humanoid robots, is gaining attention, with expectations for significant advancements in 2026 [4] - Major robotics companies are collaborating with the 2026 CCTV Spring Festival Gala, showcasing the potential for humanoid robots to perform real-world tasks [4] - Breakthroughs in action models for humanoid robots are predicted to enhance their operational capabilities in unstructured environments [5] Group 4 - The commercial space sector is identified as a significant growth area, with expectations for a pivotal moment in 2026 when SpaceX is projected to go public [6] - Chinese commercial space companies are likely to successfully implement rocket recovery operations, marking a significant milestone for the industry [6] - The low Earth orbit satellite internet sector is entering a phase of industrial explosion, with a ten-year growth cycle anticipated [6][7] Group 5 - The rocket segment is under scrutiny due to its potential to address capacity bottlenecks and high costs, with a focus on companies that can successfully reduce costs [7] - The satellite operations market is viewed as a scarce resource, with significant potential for growth in traditional high-end markets and emerging low-altitude economies [7] - The concept of space computing, utilizing AI data centers in space, is being explored as a revolutionary approach to energy and computational efficiency [7] Group 6 - The 14th Five-Year Plan emphasizes the importance of quantum technology, biomanufacturing, hydrogen energy, and other advanced fields as new economic growth points [9] - Key technological breakthroughs in areas like quantum computing and brain-machine interfaces are expected to occur by 2026, significantly impacting market confidence [9][10] - The Chinese pharmaceutical industry is poised for a breakthrough in global drug development, with several companies expected to achieve significant clinical trial results in 2026 [10] Group 7 - AI-driven drug development is anticipated to reach a validation point in 2026, showcasing the potential of AI in creating tangible medical solutions [10] - The brain-machine interface sector is expected to achieve clinical breakthroughs, particularly in flexible electrodes and signal processing chips [10] - Investment strategies are shifting towards companies that can effectively translate technological innovations into commercial success, moving away from resource accumulation [11]

Core Insights - The software sector is experiencing a significant sell-off, termed "SaaSpocalypse," with major companies like Thomson Reuters and Salesforce facing steep declines in stock prices, with the S&P 500 Software and Services Index dropping nearly 13% over five trading days and 26% from its October peak [1] - The launch of Anthropic's Claude Cowork, a general AI agent capable of interacting with user files, has raised concerns about the viability of traditional SaaS business models, as users may achieve results comparable to expensive enterprise software through natural language commands [1] - The current market panic mirrors a previous incident involving DeepSeek, which also caused a rapid sell-off in tech stocks, highlighting a pattern of investor reaction to emerging AI technologies [2][3] Market Reactions - The sell-off triggered by Cowork has been more prolonged than the DeepSeek panic, with the latter's effects dissipating within a day, while Cowork's impact has spread over a week, affecting markets globally [5] - The Cowork panic is driven by a closed-source model from a U.S. company, contrasting with DeepSeek's open-source model from China, suggesting a deeper challenge to established software companies [4] SaaS Pricing Models - Traditional SaaS companies are transitioning from fixed pricing models to usage-based pricing due to the increased efficiency and accessibility of AI, with 79 out of 500 tracked companies adopting point-based pricing, a 126% increase year-on-year [8] - Companies like Salesforce have struggled with pricing strategies, leading to a shift towards usage-based models as they face rising operational costs [8] AI Integration Challenges - Traditional software companies face resistance to price increases associated with AI integration, as seen in Microsoft's case, where customers rejected higher fees for AI features [9] - Many established firms are struggling to effectively incorporate AI into their existing products, leading to inefficiencies and a lack of user engagement [9] Emerging Trends - The rise of Vibe Coding, which allows individuals and companies to create their own tools, poses a threat to traditional software sales, as businesses may prefer to develop customized solutions rather than purchase third-party software [11] - The demand for software is shifting towards solutions that address specific, non-standard needs, indicating a potential decline in the traditional SaaS model [13]