2023年3月，北京智源人工智能研究院副院长林咏华在由界面新闻主办的2023 REAL科技大会上发表了题为《打造大模型技术的“Linux”，为AI未来十年发展打下坚实根基》的主题演讲。他指出，基础大模型是AI发展的最后一公里问题，而实现AI落地、达到产业、工业的发展要求，则需要解决这一问题。
林咏华表示：“质量成为实现AI落地的‘最后一公里’问题会比小模型更具挑战。我们致力于打造大模型的‘Linux’，让AI技术真正走进千家万户，为AI未来十年的发展奠定坚实根基。”
据悉，林咏华所提出的AI大模型“Linux”旨在为AI开发者提供一个开放、通用的平台，以满足不同领域、不同场景下的AI应用需求。该平台将有助于实现AI技术的跨领域、跨模态、跨任务整合，提高AI技术的安全性、可靠性、泛化能力，从而促进AI技术的落地和产业发展。
此外，林咏华还指出，要解决AI落地问题，需要从技术、产业、生态等多方面入手。技术层面，需要提高AI算法的性能和可靠性；产业层面，需要完善AI产业链，推动AI与实体经济深度融合；生态层面，需要构建AI生态圈，促进AI技术交流与共享。
新闻翻译：
Title:
Creating an AI Large Model “Linux” – Lin Yinghua’s Themed Speech at the 2023 REAL Technology Conference
Keywords:
– Large-scale AI model
– AI deployment
– Last mile problem
– Linux
– Industry and industrial development requirements
News content:
On March 2023, during the 2023 REAL Technology Conference, Dr. Lin Yinghua, the副院长 of Beijing Academy of Artificial Intelligence, gave a themed speech titled “Creating an AI Large Model ‘Linux’ for the Future of AI Ten-Year Development.” He pointed out that the last mile problem of large-scale AI model is more challenging than small models, and solving this problem is necessary for AI technology to truly enter households and contribute to the development of AI technology.
Lin Yinghua said: “Quality becomes the last mile problem for AI deployment. We are committed to creating an AI Large Model ‘Linux’ that truly enters households and meets the AI application needs of different fields and scenarios.”
It is reported that the AI Large Model ‘Linux’ proposed by Lin Yinghua aims to provide an open and universal platform for AI developers, satisfying the AI application needs of different fields, modalities, and tasks. This platform will help to integrate cross-domain, cross-modal, and cross-task AI technology, improve AI technology’s security, reliability, and generalization ability, thus promoting the deployment and development of AI technology.
In addition, Lin Yinghua also pointed out that to solve the AI deployment problem, it is necessary to take actions from technology, industry, and ecology. From a technical perspective, it is necessary to improve AI algorithms’ performance and reliability; from an industrial perspective, it is necessary to improve the AI industry chain, promote AI to deeply integrate with the real economy, and from an ecological perspective, it is necessary to build an AI ecosystem, promote AI technology exchange and sharing.
【来源】https://www.jiemian.com/article/10288920.html
Views: 1