美国国家工程院外籍院士、资深科技专家沈向洋在最近的演讲中，描绘了通用大模型对各行业深远影响的蓝图。他指出，随着技术的不断进步，大模型将逐渐渗透并革新所有垂直行业，模型的参数规模可能将达到前所未有的万卡乃至上万亿级别，以实现真正的通用性。
沈向洋院士强调，未来的关注点可能会转向个性化个人大模型的开发。他认为，这些模型能够根据个人的需求和习惯进行定制，从而提供更为精准的服务。将个性化参数与云计算和端设备的结合，不仅将极大地提升用户体验，同时也预示着一个充满潜力的新领域即将诞生。
他指出，这种结合将使得大模型在保持高效运行的同时，兼顾隐私保护和个性化需求，为人工智能与人类交互开辟新的可能性。沈向洋的见解为业界提供了一个全新的思考方向，预示着人机关系将在通用大模型的推动下，进入一个更加智能、个性化的时代。
来源：财联社
英语如下：
**News Title:**
Keywords: Shen Xiangyang, Large Models, Personalization
**News Content:**
Title: Academician Shen Xiangyang Envisions the Future of General-Purpose Large Models: Redefining Human-Machine Relationships and Focusing on Personalization
Academician Shen Xiangyang, a foreign member of the National Academy of Engineering in the United States and a renowned tech expert, recently outlined the transformative impact of general-purpose large models across various industries. He pointed out that with ongoing technological advancements, these models will gradually infiltrate and revolutionize all vertical sectors, potentially reaching an unprecedented scale of millions or even trillions of parameters to achieve genuine universality.
Shen Xiangyang emphasized that future attention might shift towards the development of personalized individual large models. He believes that these models can be tailored according to individual needs and habits, providing more precise services. The integration of personalized parameters with cloud computing and edge devices not only enhances user experience significantly but also signals the emergence of a promising new domain.
He highlighted that this combination enables large models to maintain efficiency while addressing privacy protection and personalized requirements, opening up new possibilities for human-AI interactions. Shen Xiangyang’s perspective offers the industry a fresh perspective, suggesting that human-machine relationships, driven by general-purpose large models, will enter a more intelligent and personalized era.
**Source:** Caixin Global
【来源】https://www.cls.cn/detail/1628008
Views: 3