Core Viewpoint - Lenovo has unveiled its "Trusted Hybrid Computing Solution" in collaboration with Volcano Engine, enhancing the security capabilities of its Tianxi personal superintelligence from endpoints to the cloud, ensuring safety both online and offline [1][2]. Group 1: AI Development and Security - The rapid development of large model technology and the emergence of DeepSeek have accelerated the growth of artificial intelligence in China, posing significant challenges to data security and personal privacy protection [1][2]. - Lenovo's hybrid AI technology path, proposed in 2023, extends from personal trusted computing devices to secure personal clouds, forming a privacy-enhanced computing and full-link protection technology system [2][3]. - The Tianxi personal superintelligence is designed to ensure absolute data security during transmission and processing between endpoints and the cloud [2][3]. Group 2: Technological Features and Capabilities - The Trusted Hybrid Computing Solution is based on Lenovo Personal Cloud 1.0, featuring core technologies such as Trusted Execution Environment (TEE), trusted transmission, and endpoint-cloud mutual trust [3]. - Tianxi personal superintelligence is equipped with a 720 billion parameter large language model and offers users a dedicated memory space starting from 100GB, combined with RAG technology to create an expandable knowledge base [3][4]. - Tianxi possesses multiple interaction capabilities, including voice, text, and visual, along with full-time memory and autonomous planning execution, redefining the smart interaction experience [4]. Group 3: Market Impact and Future Prospects - Since the launch of the world's first true AI PC in May last year, Lenovo has sold over one million units, with Tianxi's daily active users growing exponentially and achieving an advanced customer net promoter score [5]. - Lenovo aims to empower individuals and industries through AI, facilitating the rapid development of China's AI industry on a global scale [5].

杨元庆认为，无论是OpenAI还是DeepSeek都是属于公有人工智能，在问答的过程中，用户必然会贡献 或者暴露你个人的数据，包括用户非常不想暴露的个人隐私数据。我相信大多数人和大多数企业不希望 暴露隐私数据，但是如果不投喂真实的数据就得不到精准的回答，这是一对矛盾。更好的解决方案就是 利用个人可信的计算，通过用户自己的设备的计算能力来对私有数据进行学习并加以推理。这也是为什 么我们开发AI PC，推出个人超级智能体的原因。 联想集团高级副总裁兼首席技术官Tolga Kurtoglu博士表示，数据安全和隐私保护在打造"超级智能 体"当中非常重要。处理数据首先是保密的，计算还有数据的处理，尤其是敏感信息需要在设备上本地 处理，应当有技术可以把这些信息去敏，然后在不同设备当中进行转移。如果需要发送到云端，也应非 常注重并且有不同的数据安全和隐私保护措施来更好处理数据、转移数据。 杨元庆补充，企业员工也需要用公有的AI，基于企业的数据，无论是财务的数据、产供销的数据、客 户的数据、员工的数据来帮助企业更精准、更有效率的决策，优化流程。保护企业数据，保护个人隐 私，保护企业知识产权恰恰是现在推广企业智能的一个主要出发 ...