Core Insights - The global computing infrastructure is entering a new round of "computing efficiency" strategic competition, with SenseTime's large device division leading in AI infrastructure and proposing a Chinese-style "computing-energy" collaborative construction paradigm [2][4] Group 1: Company Developments - SenseTime's large device division plans to launch a self-controlled "computing-energy collaborative intelligent scheduling platform" by July 2025, based on its long-term operational practices at its AIDC [2][4] - The platform integrates computing and electricity data, achieving significant economic and social benefits in its actual operation [2][4] Group 2: Technological Innovations - SenseTime's self-developed energy model utilizes a multi-modal MoE architecture, trained on vast industry knowledge texts, structured energy data, and computing monitoring indicators [2][4] - The model can predict computing load trends in advance and dynamically solve across systems by considering factors such as electricity price signals, green electricity ratios, energy storage status, and grid demand [2][4] Group 3: Performance Metrics - The energy demand prediction accuracy of the model has reached over 88%, with decision-making accuracy at over 93% [2][4] - Continuous optimization of algorithms and energy storage devices is expected to improve prediction accuracy to 90%-95% and decision-making accuracy to over 95%, establishing industry-leading standards [2][4]

Core Insights - The AI video generation industry is undergoing a paradigm shift, moving from a focus on single-frame realism to multi-episode consistency and sustainable creation [1] - The key challenge for the industry by the second half of 2025 will be whether content can be sustained over multiple episodes, emphasizing character consistency and world coherence [1] Group 1: Industry Trends - AI video generation technology has transitioned from "can it move" to "does it move like it should," highlighting the importance of continuity in storytelling [1] - The competition in AI video is shifting from generation capability to sustainable creation ability, indicating a deeper industry consensus [1] - The introduction of tools like Seko by SenseTime reflects a move towards integrated content production infrastructure, focusing on multi-episode creation [4] Group 2: Company Developments - SenseTime's Seko platform has attracted over 200,000 creators, including numerous short drama studios, with some works successfully launched on mainstream platforms [5] - Seko supports a combination of self-developed and third-party models, ensuring consistency across different models, which is crucial for building sustainable IP assets [5] - The Seko platform's recent version emphasizes multi-episode creation, addressing the bottleneck of commercializing AI short dramas [4] Group 3: User Experience and Market Dynamics - The competition among platforms is shifting from model parameters to user experience and workflow integration, aiming to simplify the creation process for non-technical users [6] - SenseTime's approach includes a dual-track design that caters to both low-frequency and high-frequency users, allowing for easy content generation while maintaining control for advanced users [6] - The rapid growth of creators on the Seko platform, from 100,000 to over 200,000 in a short period, indicates a shift from consumer to professional use [6] Group 4: Cost Efficiency and Future Outlook - Seko has achieved a 50% reduction in inference costs through model optimization and adaptation to domestic chips, making high-quality video production more accessible [7] - Despite challenges in copyright, content safety, and IP protection, SenseTime aims to enhance creator tools rather than create a closed content platform [7] - The ultimate goal of the technology is to enhance human creativity rather than replace it, positioning Seko as a supportive tool for creators [7]

Group 1 - The AI short drama and comic drama sector is experiencing explosive growth, with most current AI video generation products only supporting single short episodes or clips [1] - SenseTime officially launched its product release week on December 15, introducing several innovative and practical AI products, including the core product Seko 2.0, which is the industry's first integrated multi-episode generation intelligent agent [1] - Seko IDX addresses the challenge of character consistency across multiple episodes and scenes by introducing "negative reference images" during the high-noise phase of the diffusion model, ensuring stability while responding to various expressions, postures, and scene commands [1] Group 2 - Traditional workflows require over three months for a team to complete 50 episodes, but with Seko 2.0, the production cycle can be shortened by 80% to 90% [1] - SenseTime is tackling the high token demand and cost issues in AI short and comic drama multi-episode creation, where a 5-second video requires nearly 100,000 tokens, and the total token demand for 10-20 scenes can reach millions [1] - SenseTime has made two core technological breakthroughs to address cost challenges: the innovative Phased DMD distillation technology and the open-source real-time video generation inference framework LightX2V, which can generate a 5-second video in under 5 seconds on consumer-grade graphics cards [2]

Core Viewpoint - SenseTime has launched Seko 2.0, an integrated creative and multi-episode generation AI model, leveraging its self-developed Seko series models for image and video generation [1] Group 1: Product Development - Seko 2.0 is built on the SenseTime's self-developed Seko series models, including SekoIDX and SekoTalk, which form the technological foundation for multi-modal generation [1] - The Seko series models are compatible with domestic AI chips from Cambricon [1] Group 2: Strategic Partnerships - SenseTime has established a strategic partnership with Cambricon in October, focusing on the joint optimization of software and hardware [1] - The collaboration aims to enhance core model capabilities, improve computing power utilization, and increase cost efficiency [1]

这些技术推动大模型从"AIforX"（为某领域提供工具）向"AIinX"（深度嵌入业务流程）演进，为构建 能与物理环境持续交互、自我演进的智能体奠定基础。 【环球网科技报道 记者 张阳】2025年12月9日，由商汤科技与香港科技园公司联合主办的"2025商汤科 技AI论坛"在香港科学园举行。以"模型智未来"为主题，论坛聚焦大模型演进、多模态融合、具身智能 落地及AI驱动的产业范式转移，吸引全球AI研究者、企业决策者与生态伙伴共议技术拐点与商业路 径。 从感知到具身： AI 进入 " 物理世界交互 " 新阶段 商汤科技CEO徐立在开场致辞中指出："我们正经历史上最大规模的技术浪潮——AI已从感知走向生 成，从云端走向端侧，并加速迈向具身智能与世界模型。"作为一家根植香港、服务全球的AI原生企 业，商汤将持续连接国家AI战略与全球创新网络。 香港科技园CEO黄秉修则强调，商汤的成长印证了香港具备孕育世界级科技企业的土壤，"其从初创到 全球领先的跃迁，正是我们支持硬科技创业的最佳案例。" 多模态大模型进入 " 价值闭环 " 攻坚期 商汤联合创始人、首席科学家林达华教授认为：经过三年爆发式发展，行业站在新十字路口。" ...

Core Insights - The article discusses the competitive landscape of video generation models, highlighting the advancements made by various tech companies, including Google, Runway, and Kuaishou, while questioning the readiness of these models as productivity tools [2][9] - SenseTime's Seko 2.0 version is introduced as a significant advancement, enabling AI short drama creation with minimal human input, effectively allowing a single person to manage the production [2][4][7] Group 1: Industry Developments - Major tech companies are racing to release enhanced versions of video generation models before the end of the year, with Google launching Veo 3.1 and Runway introducing Gen-4.5 [2] - SenseTime's Seko 2.0 has been successfully deployed in over a hundred short drama studios, showcasing its capability to generate scripts, storyboards, and videos rapidly [7][9] Group 2: Technical Challenges - The article outlines the "impossible triangle" of video generation, where efficiency, cost, and quality are at odds, making it difficult for AI video generation models to meet commercial demands [11][13] - Current models, even at the Sora 2 level, require several minutes to generate just 10 seconds of video, which hampers rapid iteration and real-time feedback essential for industrial production [11][12] Group 3: Innovations in Video Generation - SenseTime's LightX2V framework is highlighted as a breakthrough in real-time video generation, achieving generation times of under 5 seconds for 5-second videos, significantly faster than current industry standards [16][17] - The framework employs Phased DMD technology, which enhances video quality and consistency while maintaining high generation speeds [19][20] Group 4: Engineering and Optimization - LightX2V incorporates a comprehensive optimization strategy across five dimensions: model, scheduling, computation, storage, and communication, enabling low-cost and real-time video generation [31][32] - The framework's architecture allows for efficient use of consumer-grade GPUs, achieving real-time generation capabilities with a memory requirement of less than 8GB [36][37] Group 5: Domestic Chip Adaptation - SenseTime's Seko 2.0 has achieved full compatibility with domestic AI chips, allowing for a cost-effective alternative to NVIDIA chips while maintaining comparable video quality [39][40] - The strategic support for domestic AI ecosystems is emphasized, marking a significant step for China's AI industry in achieving core technological independence [42]

Market Overview - Major Asia-Pacific stock indices closed mostly lower, with the Nikkei 225 and KOSPI indices both dropping over 1% [1] - A-shares experienced fluctuations, with the Shanghai Composite Index and Shenzhen Component Index both declining, while the Hang Seng Index fell significantly [2] - The total trading volume in the Shanghai and Shenzhen markets was 1.79 trillion yuan, a decrease of 324.6 billion yuan from the previous trading day [2] Sector Performance - Consumer stocks showed resilience, particularly in retail and food & beverage sectors, with notable performances from liquor stocks like Zhongrui Co. and Kweichow Moutai [2] - The insurance sector surged over 4%, led by companies such as China Ping An, which rose nearly 5% to a new four-year high [2][4] - The commercial aerospace sector also strengthened, with stocks like Hualing Cable and Zhongchao Holdings achieving three consecutive trading limits [4] Commodity Market - Gold and silver prices saw strong increases, with spot gold rising over 1% to reach $4,343 per ounce, and silver increasing by over 2% [7][8] - Platinum prices have surged over 100% this year, with analysts predicting further increases in the coming year [6][9] - Analysts expect platinum prices to rise significantly, with projections for 2026 ranging from $2,170 to $2,300 per ounce, indicating a shift towards undervalued precious metals [10]

Core Insights - SenseTime Technology officially launched Seko 2.0, the industry's first multi-episode generative intelligent agent, leveraging its technological accumulation in generative AI and multimodal interaction [1] - The Seko series models, including SekoIDX and SekoTalk, provide a robust technical foundation for image and video generation, showcasing significant advantages in consistency for multi-episode video generation [1] - The collaboration with Cambricon (688256.SH) marks a key advancement in supporting AIGC core scenarios, facilitating a critical leap from language to multimodal capabilities [1] Group 1 - The Seko series models have been adapted to domestic AI chips, enhancing the support for visual content innovation and development within the domestic AI ecosystem [1] - The LightX2V framework is designed with a highly compatible domestic adaptation plugin model, currently supporting multiple domestic chips, including Cambricon [1] - Innovations such as low-bit quantization, compressed communication, and sparse attention mechanisms have been integrated into the Seko series models, resulting in over three times improvement in inference performance [1] Group 2 - SenseTime and Cambricon have established a strategic partnership to optimize software and hardware jointly, aiming to create an open and win-win industrial ecosystem [2] - The collaboration focuses on continuous optimization of core model capabilities, enhancing overall efficiency and response speed for multimodal generation [2] - Efforts will be made to improve computing resource utilization and cost efficiency through operator fusion and automatic tuning, allowing more enterprises to access high-performance multimodal capabilities at lower costs [2] Group 3 - The partnership aims to foster the prosperity and development of the domestic AI application ecosystem, creating more efficient and user-friendly tiered product systems [3] - The goal is to build a more open and friendly tool and ecosystem for developers, stimulating innovation in cutting-edge applications [3]

Core Viewpoint - SenseTime officially launched Seko 2.0, the industry's first multi-episode generative intelligence agent, leveraging its self-developed Seko series models [1] Group 1: Product Development - The Seko series models have completed adaptation to the domestic AI chip Cambricon, marking a significant advancement in supporting AIGC core scenarios from language to multi-modal capabilities [1] - Following the adaptation, SenseTime and Cambricon will further collaborate on deep optimization across multiple directions [1]

Group 1 - SenseTime officially launched Seko 2.0, the industry's first multi-episode generative agent, showcasing significant advantages in consistency for multi-episode video generation [1] - The Seko series models, including SekoIDX and SekoTalk, are built on SenseTime's proprietary technology, which has been adapted to support domestic AI chips from Cambricon, marking a key leap from language to multi-modal capabilities [1] - The LightX2V framework is designed with a highly compatible domestic adaptation plugin model, currently supporting multiple domestic chips, including Cambricon, enhancing the performance of the Seko series models [1] Group 2 - In October, SenseTime and Cambricon established a strategic partnership to optimize software and hardware jointly, facilitating a collaborative innovation between domestic large models and computing power [2] - The partnership aims to continuously optimize core model capabilities, enhance computing efficiency, and reduce resource consumption, allowing more enterprises to access high-performance multi-modal capabilities at lower costs [2] - The collaboration will also focus on improving large-scale parallel processing capabilities and developing a more flexible resource management mechanism to ensure stable model operation across diverse environments [2] Group 3 - The deep collaboration between SenseTime and Cambricon is expected to significantly enhance model efficiency, resource utilization, and cross-hardware compatibility, lowering the barriers to using multi-modal AI [3] - The partnership aims to foster a thriving domestic AI application ecosystem, creating more efficient and user-friendly product systems while providing developers with open and friendly tools [3]