Core Insights - The conference "Going Global with Craftsmanship" aims to provide a platform for companies to explore globalization strategies and opportunities in various sectors such as consumption, technology, e-commerce, finance, and new energy [1] - The event features discussions on the collaboration between products, technology, and ecosystems to enhance sustainable global expansion capabilities for enterprises [1] Group 1: AI Empowerment in Global Expansion - The presentation by SenseTime's executive director focuses on how generative AI can enhance operational, marketing, and service aspects of companies going global [3] - SenseTime is recognized as the world's first publicly listed pure AI company, emphasizing its commitment to developing industry-specific tools for easier implementation [3] - The discussion highlights three main areas of AI application: digital human live streaming, content generation, and AR product effects [4][6][8] Group 2: Digital Human Live Streaming - Digital human live streaming balances efficiency and cost, providing interactive experiences similar to real human hosts while being able to respond to audience queries in real-time [4] - This technology allows for quick adaptation to various scenarios and creates a full marketing loop, generating promotional content and managing online reputation [4][6] - A single digital human can operate across multiple platforms, replicating the style of popular influencers, thus enabling cost-effective scaling [5] Group 3: Content Generation with AIGC - AIGC (AI-Generated Content) streamlines the marketing process by integrating content creation, distribution, and optimization, significantly reducing costs [6] - It enables the generation of high-quality advertising materials from simple product descriptions, facilitating efficient marketing across different platforms [6][7] - The technology can lower operational costs by over 50% compared to traditional marketing methods [7] Group 4: AR Technology for Enhanced Experience - Collaboration with Zhejiang University has led to the development of AR applications that improve product interaction and customer experience [8] - Virtual try-ons for cosmetics and accessories enhance user engagement and efficiency, reducing the need for physical trials [8] - The technology extends to various sectors, including e-commerce, supply chain management, and smart home applications, showcasing its versatility [8][9] Conclusion - AI is viewed as a revolutionary tool that transforms the cost-efficiency-experience triangle in global marketing strategies, with ongoing advancements expected to unlock further value across industries [10]

Core Viewpoint - The development of multi-modal information perception and processing capabilities is essential for achieving Artificial General Intelligence (AGI), marking a significant transition from language models to AGI [1][3]. Group 1: SenseNova V6.5 Model Upgrade - SenseNova V6.5 introduces three major breakthroughs: enhanced reasoning capabilities, improved efficiency with a cost-performance ratio increased by over 300%, and advanced data analysis leading to end-to-end scenario implementation [3][4]. - The model's multi-modal reasoning and interaction capabilities have significantly improved, surpassing competitors like Gemini 2.5 Pro and Claude 4-sonnet in text reasoning and multi-modal interaction [4][5]. - The new architecture promotes early cross-modal fusion, resulting in a 20% increase in pre-training throughput, a 40% boost in reinforcement learning efficiency, and a 35% improvement in reasoning throughput [5]. Group 2: Application of Multi-Modal Capabilities - The upgraded SenseNova V6.5 enables the "Xiaohuanxiong" AI assistant to handle complex multi-modal inputs, providing in-depth analysis and professional visualization outputs, thus transforming AI from a productivity tool to a true productivity driver [6][8]. - Xiaohuanxiong achieves near 100% accuracy in tasks such as time series calculations, data matching, mathematical computations, and anomaly detection, positioning it at the international benchmark level [6][10]. - The AI assistant can simplify complex data inputs, such as Excel sheets with merged cells and nested tables, and generate comprehensive analysis reports [10][12]. Group 3: Industry Impact and User Engagement - The Xiaohuanxiong assistant has been deployed in various sectors, including education and finance, with over 10 million users benefiting from its capabilities [15]. - In the education sector, it has improved student learning efficiency by 15-30% and reduced academic anxiety by 40% across more than 500 institutions [13]. - The financial version of Xiaohuanxiong offers solutions for knowledge assistance, intelligent querying, and multi-modal claims processing, establishing a new paradigm for human-machine collaboration in decision-making [14].

Summary of Key Points Core Viewpoint - The report highlights the top short-selling stocks in the market, indicating significant short-selling activity and potential investor sentiment towards these companies. Group 1: Top Short-Selling Ratios - JD Health (86618) has the highest short-selling ratio at 100.00% [1][2] - Hang Seng Bank (80011) follows with a short-selling ratio of 88.44% [1][2] - SenseTime (80020) has a short-selling ratio of 76.25% [1][2] Group 2: Top Short-Selling Amounts - Xiaomi Group (01810) leads in short-selling amount with 2.209 billion [1][2] - Tencent Holdings (00700) has a short-selling amount of 0.955 billion [1][2] - WuXi AppTec (02359) reports a short-selling amount of 0.698 billion [1][2] Group 3: Top Short-Selling Deviations - Hang Seng Bank (80011) has the highest deviation value at 48.82% [1][2] - JD Health (86618) follows with a deviation value of 45.78% [1][2] - Uni-President China (00220) has a deviation value of 33.70% [1][2]

Group 1: AI Hardware Trends - The integration of AI technology into daily life is accelerating, with consumer-grade AI hardware emerging as the first carriers of this integration [1] - Various companies are launching diverse smart hardware products, including smart glasses, AI toys, and smart office devices, targeting specific verticals like elderly care and emotional companionship [1] Group 2: Smart Glasses Market - Smart glasses have become a focal point at the World Artificial Intelligence Conference (WAIC), with brands like XREAL showcasing their flagship product, XREAL One Pro, which features a wide field of view and low latency [2][3] - Other notable players include Rokid, which introduced Rokid Glasses with a teleprompter function, and Halliday, which launched Halliday Glasses that support prescription lenses and offer long battery life [3] Group 3: AI Companion Products - AI companion products are evolving beyond traditional toys, with offerings like the Fuzozo robot from SenseTime, which features realistic digital avatars and enhanced interaction capabilities [4] - Looi, a desktop robot, allows users to engage with it through emotional recognition and interactive games, while also providing development tools for customization [5] Group 4: Focus on Office and Education - AI hardware is increasingly being applied in office and educational settings, with companies like Outermost introducing products like TicNote, which combines hardware and software for various practical applications [6] - NetEase Youdao showcased the AI-powered learning tool SpaceOne, designed to enhance students' problem-solving skills and reduce dependency on standard answers [6] Group 5: Industry Challenges - Despite the excitement surrounding AI hardware, challenges remain, such as limitations in computational power and energy consumption, which hinder the full potential of AI functionalities [7] - The industry faces the core challenge of promoting deep integration between AI technology and hardware to create next-generation smart devices with continuous evolution capabilities [7]

Core Insights - The essence of artificial intelligence (AI) is to create a form of genuine intelligence that can autonomously interact with the real world, which is the ultimate goal of intelligence [1] - The rapid evolution of large models, particularly language models, is seen as a stepping stone towards achieving AGI (Artificial General Intelligence), with a necessary focus on multimodal capabilities for real-world applications [1][2] Company Developments - SenseTime has officially launched the "Riri Xin" V6.5 "Awakening" world model and the "Wuneng" embodied intelligence platform during the WAIC [1] - The company has been a pioneer in multimodal integration, demonstrating that multimodal models outperform pure language models in language tasks after effective training [2] - The latest version, "Riri Xin" 6.5, has achieved advanced performance in both pure language and text tasks, showcasing the maturity of SenseTime's technology in this area [2] Industry Trends - The rise of ChatGPT has highlighted a new era in AI technology, presenting opportunities for companies like SenseTime to leverage this wave of transformation to create significant impact [3] - The shift from AI 1.0, which focused on specialized tasks, to general AI models that are more autonomous and versatile is a key development in the industry [3] - The future of software development is expected to become more accessible, allowing non-experts to create software simply by expressing their needs, which could reshape industry dynamics [3][4] Technological Advancements - The development of multimodal models is progressing through three critical stages, with the final goal being the connection between digital and physical spaces to achieve AGI [5] - SenseTime's experience in computer vision and collaboration with hardware companies has positioned it well to enhance its embodied intelligence platform [6] - The integration of world models with multimodal training data has proven effective in training autonomous driving modules, significantly improving efficiency compared to relying solely on real-world data [6] Strategic Focus - SenseTime emphasizes aligning research and development with its commercial vision, ensuring that scientific advancements translate into business value [6] - The company prioritizes projects that can achieve commercial viability, avoiding areas that do not align with its business goals [6] - Investments in embodied intelligence and foundational models are interconnected, allowing for a more efficient allocation of resources [6]

Core Insights - The core issue in end-to-end autonomous driving is the need for massive data collection and the ability to cover high-risk scenarios, which poses a data bottleneck for training models [2][4] Group 1: Company Developments - SenseTime's "Jueying Kaiwu" world model is the first interactive generative world model product platform in the assisted driving sector, aimed at addressing data collection challenges [4] - The platform can generate millions of scene data and create real-time interactive training environments, significantly enhancing the assisted driving industry [4] - The efficiency of the "Jueying Kaiwu" model is notable, as it can generate data equivalent to that collected by 500 mass-produced vehicles using just one A100 GPU [4] Group 2: Industry Challenges - The lack of training data is a significant barrier to the widespread adoption of intelligent robots, with leading companies only producing limited real-world data [5] - The growth of visual data generation is lagging behind computational power, leading to a mismatch in model data requirements [5][6] - The need for a large-scale 4D spatial reconstruction capability is essential for creating realistic training scenarios, including high-risk collision scenarios [7] Group 3: Future Implications - The introduction of world models can enable autonomous evolution of driving behavior by simulating various real-world changes and generating multi-modal data [7] - The relationship between humans and AI must be carefully managed to maintain human uniqueness in the era of human-machine coexistence [8][9] - Defining rules and values for AI interactions is crucial for ensuring that robots develop intelligence within acceptable boundaries [9]

Group 1 - Companies such as Midea, JD.com, Tencent, and SenseTime are showcasing their developments in embodied intelligence and humanoid robots, indicating a need for further enhancement of AI hardware and software [1] - The upgrade of AI hardware is driving high demand for high-end PCBs, with analysis suggesting that only a few global manufacturers possess stable mass production capabilities, leading to a short-term inability to meet demand due to technical barriers [2] - A seminar on the standardization of large model integrated machines is imminent, with institutions stating that the demand for large model integrated machines in China is set to explode, potentially opening up a market worth hundreds of billions [3] Group 2 - Meta is developing two third-generation smart glasses in collaboration with Ray-Ban, with reports indicating that these products may be released ahead of schedule in September [4] - Saiwei Electronics has seen the National Integrated Circuit Fund reduce its stake in the company by 1.06% [5] - Southern Road Machinery has maintained its focus on the engineering mixing field, with no changes to its main business despite experiencing a seven-day stock surge [6] - Fangbang Co., Ltd. reported that sales revenue from its copper-strippable products accounted for less than 0.3% of its main business income in the first half of the year [7] - Tibet Tourism has confirmed that it has not engaged in business cooperation with entities related to the downstream hydropower project of the Yarlung Tsangpo River [8] - Baotou Steel has been confirmed as the winning bidder for the Yarlung Tsangpo rare earth alloy steel project [9] - Liyuan Technology's actual controller, Shen Wanzhong, has been sentenced to one year in prison for the crime of illegally disclosing important information [10] - WuXi AppTec plans to adjust the upper limit for its share repurchase price to no more than 114.15 yuan per share [11]

Group 1 - Sense, navigation, and interaction are the three core capabilities of embodied intelligence, with Sense being the foundation for machines to explore the real world [2] - The "Wuneng" embodied intelligence platform by SenseTime integrates advanced visual AI technology to provide recognition and understanding capabilities for various hardware terminals [2] - Navigation is described as the "skeleton" for machines to act in the real world, with SenseTime's technology enabling precise path planning and navigation for robots and other devices [2] Group 2 - The "Wuneng" platform allows robots to interact with the real world, showcasing capabilities such as warmth, depth, long memory, and stability [2] - SenseTime, in collaboration with various domestic partners, launched the "SenseTime Computing Power Mall" to provide flexible and autonomous domestic computing power options [3] - The "SenseTime Computing Power Mall" aims to lower the barriers for AI applications and promote the independent and controllable development of China's AI industry [3]

经济观察网 据商汤科技微信公众号消息，7月27日，商汤科技发布「悟能」具身智能平台，基于方舟视 觉感知、大装置、日日新大模型等多种技术的长期积累，商汤「悟能」具身智能平台以商汤"开悟"世界 模型为核心引擎，依托商汤大装置提供强大的端侧和云侧算力支持，能够为机器人、智能设备赋予强大 的感知、视觉导航及多模态交互能力。 商汤科技董事长兼首席执行官徐立表示，世界模型和具身AI的结合，将开启AI发展的下一个阶段，完 成由"工具"向"人"的跃迁，加速迈向AGI时代。商汤希望「悟能」具身智能平台能够帮助各类具身智能 企业，帮助他们完成和现实世界交互的梦想。 ...

Core Insights - The rise of large language models (LLMs) marks a significant leap in AI technology, but achieving Artificial General Intelligence (AGI) requires more than just text understanding and generation [1] - The future of AI development lies in the integration of multimodal information and interaction with the physical world, with a shift towards multimodal models expected to accelerate [1][2] - The realization of AGI necessitates long-term technological accumulation and iterative scene development, overcoming key bottlenecks such as spatial perception and data scarcity [2][8] Multimodal Development - The evolution of large models is transitioning from single-language models to native multimodal architectures, which integrate various types of information during the pre-training process [4][5] - Current multimodal models need to extend from understanding to thinking, incorporating both logical and visual thinking processes [4][5] - Domestic companies are expected to adopt multimodal models comprehensively by the second half of 2025, moving away from standalone language models [5] Challenges in Achieving AGI - Key challenges include the generalization of reasoning capabilities from narrow domains to complex real-life scenarios, as well as the current limitations in spatial perception of multimodal models [2][7] - The development of agents, seen as crucial for AI's real-world application, faces significant gaps in understanding complex conditions and specific industry needs [6][7] - The ability of agents to effectively solve problems in real scenarios is essential for their perceived value and reliability [6] Bottlenecks in Embodied Intelligence - Embodied intelligence must bridge the gap between digital and physical spaces, with current data acquisition methods relying heavily on limited robotic operations [8] - The data throughput for embodied intelligence is significantly lower than that available from the internet, creating a challenge for effective development [8] - To advance embodied intelligence, leveraging prior knowledge and multimodal data from the internet is necessary, as relying solely on real-world data is insufficient [8]