is there a bubble in agi? han xinyi, wang xiaochuan, and yan junjie responded at the bund conference
2024-09-06
한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina
cover news reporter zhang yuexi
on september 5, at the "industry innovation roundtable" session of the main forum of the 2024 inclusion·bund conference, industry leaders including han xinyi, president and cfo of ant group, wang xiaochuan, founder and ceo of baichuan intelligence, and yan junjie, founder and ceo of minimax, discussed topics of concern to the industry, such as the development process of large models, breakthroughs in the next cycle, and the path to industry realization. the guests believed that application breakthroughs are the key to the development of china's general artificial intelligence technology (agi).
2024 inclusion·the bund summit main forum "industry innovation roundtable" session
the guests reached a consensus that the development of large models needs to be viewed over a longer period of time. in the past two years, large model technology has made breakthrough progress. currently, the industry is at a node in development. large models are moving from "competing parameters" to "competing applications". next, the industry faces two problems: one is how to reduce its inference cost, thereby promoting large-scale landing applications. the other is how to use applications to drive the development of technology.
han xinyi believes that the development and application of ai technology is a continuous iterative process, which requires discovering the difficulties of the technology in the application to make the direction of technological development more focused.
wang xiaochuan denied the "agi bubble" claim, and he believed that the development of big models was in line with industry expectations. he pointed out that in order to push big models to the next stage of development, the key lies in achieving two major integrations: one is to combine knowledge search with big models, and the other is to combine big models with reinforcement.
big models are considered to be the only way to achieve true general artificial intelligence. wang xiaochuan believes that when talking about agi, we can refer to the standard of "human", that is, thinking, knowledge, and providing service capabilities like humans, and when it comes to scenarios, it is an assistant like a human.
yan junjie feels that it will take patience to achieve such a standard. at present, he is more concerned about expanding the boundaries of ai through technical optimization. he said that if the model error rate can be reduced to single digits, then the model will be trustworthy, and ai may also change from an assistant to a kind of productivity. this change will also bring about truly popular products and bring about essential changes to the industry.
at the roundtable, the guests also shared the big model application practices in their companies.
han xinyi said that the goal of large-scale model products can be summarized as: easy to use, useful and affordable. for users, this means a new and more intuitive interactive experience; for merchants, it means being able to understand user needs more deeply and provide more structured and personalized services. this kind of service is a delivery result, and its transformation is not only reflected in user interaction, but also a deep-level change in the supply side. alipay is constantly exploring how to integrate artificial intelligence technology more naturally into people's daily lives and business operations, so as to promote innovation and transformation of business models.
han xinyi also revealed that ant will release three products at the bund summit: ai life butler "zhi xiaobao", ai financial butler "ant xiaocai" and ai health butler, among which zhi xiaobao is an independent app. he expects that these three ai butlers will provide good services to users in three fields, and at the same time connect the open ecosystem in the form of intelligent entities to find a business model in the new business environment.
baichuan intelligence focuses on investing in the field of large-scale medical ai, starting from the two entry points of pediatricians and general practitioners, hoping to achieve low cost and accessibility of medical care next year, alleviate anxiety for users, and reduce hospital runs.
