【环球网科技综合报道】9月10日-14日，以"数智领航，服贸焕新"为主题的2025年中国国际服务贸易交易会（简称"服贸会"）在北京举办。当前全球服务贸 易正加速向数字化、智能化转型，数字服务、知识密集型服务已成为全球服务贸易增长的核心引擎。作为全球重要的综合型服务贸易展会，服贸会已成为展 现全球服务贸易趋势、促进国际服务合作的核心平台。 作为国家语言服务出口基地，传神语联受邀参加全球服务贸易峰会，携多款AI创新成果参展，覆盖语言服务、中医药服务、跨境商务服务等服贸重点领 域，以"根原创"技术优势，为产业智能化升级提供可落地的解决方案。 AI 创新成果落地，多领域赋能服贸升级 本届服贸会专题展聚焦服务贸易融合发展，重点展示数字技术、人工智能等新技术在服贸领域的应用。在传神语联展位，任度大模型、传神素问中医大模 型、传神小尾巴AI翻译机、全球会客厅等产品受到观众关注。 其中，基于"根原创"任度大模型打造的传神素问中医大模型，专注于中医传承与辅助诊疗，在智能问诊、辨证分析、方剂推荐、健康管理等方面提供体系化 支撑。该模型近日通过中国信通院可信AI中医药大模型评估，获得4+级高评级，为中医文化传承与出海奠定坚实基础。 ...

Group 1 - The theme of this year's Davos Forum is "Entrepreneurship in the New Era," emphasizing the need for entrepreneurs to play a greater role and showcase more achievements, particularly in the AI industry [1] - The AI core industry in China is approaching a scale of 600 billion yuan, with over 4,500 related companies covering key aspects of the industry chain, including chips, algorithms, data, platforms, and applications [2] - The development of AI relies on three core elements: algorithms, computing power, and data, which are interdependent and drive technological progress [2] Group 2 - Chinese AI companies are encouraged to adhere to "root innovation," focusing on core algorithms and hardware systems to achieve self-control and align with national calls for original technology [3] - The company Transn has developed a complete original AI technology stack, including the zANN algorithm framework and moH mixed entropy model architecture, which reduces dependence on large computing power [4] - Transn's commitment to "root originality" has led to the launch of various products, including the RenDu dual-brain model integrated machine and models for traditional Chinese medicine and agriculture [4][5] Group 3 - The concept of "sovereign large models" proposed by Transn emphasizes training AI with data that reflects Chinese cultural values, allowing AI to understand and convey the richness of Chinese culture while serving global scenarios [7] - This transformation is not only a technological breakthrough but also relates to cultural sovereignty and technological security, highlighting the importance of mastering underlying algorithms and model architectures [7]

Core Insights - The article discusses the transformation of the intelligent language service industry, highlighting the shift from simple language conversion to knowledge understanding and application [2][3][4] - The competitive landscape of large models is evolving, with new players emerging and the industry still in its early stages, akin to the electrical era of the 1920s [2][12] - The paradox of increasing order volume but stagnant profitability in the intelligent language service sector is emphasized, indicating a need for companies to adapt their strategies [5][6] Industry Trends - The intelligent language service industry is experiencing a significant transformation, where machine translation is becoming more prevalent, yet human translators are still needed for quality assurance [4][5] - The demand for intelligent language services is growing, with a reported 30% increase in order volume for a leading company, while revenue only grew by 10% [5] - The industry is moving towards a model where companies provide foundational technologies and tools to partners rather than directly serving end customers [6][7] Data Quality vs. Quantity - The focus is shifting from "big data" to the quality of data, as high-quality data is essential for effective machine learning and AI applications [7][8] - Companies are encouraged to separate data from reasoning to enhance AI systems' adaptability and efficiency [8][9] - The value of data is increasingly recognized as being tied to its knowledge density rather than sheer volume [9][10] Future of Large Models - The current state of large models is not yet mature, and the market dynamics are still evolving, with many applications yet to be discovered [10][12] - The competition in AI is expected to focus more on foundational technology frameworks rather than just parameter size [11][12] - The future of AI services in the enterprise market is unlikely to be free, as solving business problems incurs costs [12]