p 群里人都死光了吗 ...

Core Viewpoint - Cowork, the latest product from Anthropic, is an AI coding tool designed for non-developers, integrating AI into the development process itself [1] Group 1 - Cowork is aimed at ordinary users who do not need programming knowledge, as it automates coding tasks within the workflow [2] - Users can grant Cowork access to specific folders on their computers, allowing it to perform tasks such as organizing files, writing reports, and creating spreadsheets based on screenshots [2] - The tool operates more autonomously than typical chat applications, planning and executing tasks while providing progress updates [2] Group 2 - Currently, Cowork is only compatible with macOS, utilizing a local virtual machine to read documents with fine-grained permission control [2] - This integration of cloud AI with local documents enhances both intelligence and security, creating significant potential for future applications [3] - The overseas reception of Cowork has been very positive, suggesting that domestic AI models could learn from this approach to combine cloud and terminal AI for personalized experiences [3]

Core Insights - The article highlights a significant increase in third-party API token usage, reaching a new high, which was predicted two weeks prior [3] - The domestic MiMo platform ranks third globally in terms of performance [3] Group 1 - The total API token usage reached 7.11 trillion, with a weekly increase of 547 billion [2] - The top contributors to the API token usage include Claude Opus 4.5 at 599 billion and Claude Sonnet 4.5 at 580 billion [2] - Other notable contributors include MiMo-V2 -Flash at 506 billion and Grok Code Fast 1 at 432 billion [2]

Group 1 - Nvidia has quietly revised its previous forecast for copper usage in data centers, reducing the amount of copper busbars per gigawatt from 500,000 tons to 200 tons [1] - This revision may indicate that prior expectations for copper consumption in data centers were significantly overstated [1] - Copper prices have retreated from historical highs, with a decline of up to 2% [1]

Group 1 - The company engaged in six inquiries on Qwen, including two overseas data points and four domestic ones [1] - The company plans to utilize Qwen for real-time translation during an overseas technical discussion [1] - The company identifies itself as a premium customer, consistently paying for various streaming services annually [1]

Group 1 - The article discusses the convenience of using a voice-activated system for ordering food, highlighting the speed and efficiency of the process, which can take as little as one minute [20] - It emphasizes the potential for the system to learn user preferences over time, such as favorite restaurants and dietary restrictions, thereby enhancing the user experience [20] - The integration of health data to provide personalized recommendations, such as reducing oil and salt intake, is also mentioned as a significant feature of the service [20] Group 2 - The article includes various food items available for order, with specific prices listed, such as ¥6.9 for a dessert and ¥1.5 for a sauce [10][12][15] - It notes the popularity of certain items, with recent positive reviews from customers, indicating a favorable reception in the market [13][16][17] - The pricing strategy is highlighted, showing a range of options from low-cost items to more expensive meal combinations, catering to different consumer preferences [19]

我宣布 它以后是我每日用app了！ 我先不给大家安利工作流。因为我还没上传工作文档给它。要先清洗一下关键信息。 我刚刚试了一下实时翻译！ 我直接试的中翻英！ 看了准确度，非常高！ 比我之前用X上的 实时翻译准确度高很多。 还有文档发送给我。 以后黄皮衣的现场，那不是他讲完 我三分钟后就能点评完了？nice！ 完美！直接翻译屏幕 ...

Group 1 - The article highlights the rapid rise of various individuals in the industry, indicating a dynamic and competitive environment [1] - It mentions specific names such as 阳阳, 柱柱, 威威, 蒙蒙, and 聪聪, suggesting they are gaining prominence and influence [1] - The text implies that these individuals are now involved in product management, indicating a shift in their roles and responsibilities within the industry [1]

Core Insights - The article highlights the emergence of domestic AI models, particularly focusing on Xiaomi's MiMo-V2-Flash, which has shown significant advancements in efficiency and performance [1]. Group 1: MiMo-V2-Flash Overview - MiMo-V2-Flash was officially released and open-sourced on December 17, featuring a total parameter count of 309 billion and active parameters of 150 billion, utilizing a mixture of experts (MoE) architecture [1]. - The model achieves a remarkable inference speed of 150 tokens per second, with costs reduced to $0.1 per million tokens for input and $0.3 for output [1]. Group 2: Cost Efficiency Innovations - The model employs a hybrid sliding window attention mechanism, which significantly reduces the storage of KV cache by nearly six times while maintaining long text capabilities, supporting a context window of approximately 256k [9]. - MiMo-V2-Flash integrates a three-layer MTP setup, enhancing encoding task speed by about 2.5 times and addressing GPU idle time issues in small batch on-policy reinforcement learning [10]. - The model utilizes a multi-teacher online policy distillation (MOPD) approach, requiring only 1/50th of the computational power of traditional methods to achieve peak performance, allowing for faster model iterations and self-evolution [10]. Group 3: Competitive Positioning - MiMo-V2-Flash has achieved high rankings in benchmark tests, placing among the top two in the AIME 2025 mathematics competition and GPQA-Diamond science knowledge test, with a programming capability score of 73.4% in the SWE-bench Verified test [6][11]. - The model's performance is noted to be competitive with leading models, closely approaching GPT-5-High in programming tasks [6].

Core Insights - The article discusses the competitive landscape of AI models, highlighting the performance of various models including Grok, Mimo-V2, and others, indicating a rapid evolution in the sector [2][4]. Group 1: AI Model Performance - Grok has surpassed Gemini in terms of performance, indicating a significant shift in the competitive dynamics of AI models [2]. - Mimo-V2 from Xiaomi is noted as a strong contender, ranking third in the performance metrics [2]. - The overall performance of AI models is expected to reach new highs in the upcoming week, suggesting ongoing advancements in technology [2]. Group 2: Performance Metrics - The total performance of the AI models listed amounts to 6.43 trillion (T) [4]. - Claude Sonnet 4.5 leads with a performance of 531 billion (B), followed by Grok Code Fast 1 at 413 billion (B) and MiMo-V2-Flash at 398 billion (B) [4]. - Other notable models include Gemini 3 Flash Preview at 387 billion (B) and DeepSeek V3.2 at 312 billion (B), showcasing a diverse range of capabilities among the top performers [4].