美国国家工程院外籍院士、资深科技专家沈向洋近日在一次公开演讲中指出，我们正步入一个由通用大模型主导的新时代，这些模型预计将颠覆各个垂直行业。沈向洋强调，为了实现通用模型的高性能，模型的参数规模将达到前所未有的万卡乃至上万亿级别，这在科技领域引发了深远的思考。
沈向洋院士进一步提出，未来的关注点可能将转向个性化的个人大模型。他认为，将个人化的参数与云计算及端设备的先进技术相结合，将为人工智能的发展开辟全新的可能性。这种创新模式不仅将提升用户体验，还能在数据安全和隐私保护的前提下，为用户提供更为精准的服务。
他表示，这种结合将创造出极具价值的新局面，有可能重塑人机交互的方式，推动科技与人类生活的深度融合。沈向洋的见解为人工智能领域的发展提供了新的视角，预示着一个更加个性化、智能化的未来正在逐步形成。
来源：财联社
英语如下：
**News Title:** “Academician Sheng Xiangyang Envisions: The Era of General Large Models Will Revolutionize Human-Machine Relationships, with Personalization as the New Focus”
**Keywords:** Sheng Xiangyang, Large Models, Personalization
**News Content:**
Title: Academician Sheng Xiangyang Foresees a Transformation in Human-Machine Relations with the Advent of General Large Models, Focusing on Personalization
Renowned science and technology expert, Academician Sheng Xiangyang, an overseas member of the U.S. National Academy of Engineering, recently highlighted in a public lecture that we are entering a new era dominated by general large models, which are expected to disrupt various vertical industries. Sheng emphasized that to achieve high performance in these models, their parameter scale could reach an unprecedented level of tens of billions or even trillions, sparking deep contemplation within the tech sector.
Academician Sheng further suggested that the future focus might shift towards personalized individual large models. He believes that combining personalized parameters with cutting-edge technologies in cloud computing and edge devices will unlock new possibilities for artificial intelligence development. This innovative approach not only enhances user experience but also enables more precise services while ensuring data security and privacy.
He stated that this integration could create a valuable new landscape, potentially reshaping the way humans interact with machines and fostering a deeper fusion of technology with daily life. Sheng Xiangyang’s insights offer a fresh perspective on the development of artificial intelligence, signaling the emergence of a more personalized and intelligent future.
**Source:** Cailian Press
【来源】https://www.cls.cn/detail/1628008
Views: 1