Alibaba Cloud Intelligent CTO Zhou Jingren: MaaS has become an industry standard, and future applications will be developed around models

On May 28th, the 2023 Zhongguancun Forum on the Development of Artificial Intelligence Big Models was held in Beijing. Zhou Jingren, CTO of Alibaba Cloud Intelligence Group, delivered a keynote speech titled "MaaS, A New Paradigm of AI in the Cloud Intelligence Era"

On May 28th, the 2023 Zhongguancun Forum on the Development of Artificial Intelligence Big Models was held in Beijing. Zhou Jingren, CTO of Alibaba Cloud Intelligence Group, delivered a keynote speech titled "MaaS, A New Paradigm of AI in the Cloud Intelligence Era". Zhou Jingren said, "The model centered development paradigm (MaaS) has become an industry standard, and the entire chain of future application development will be based on this concept

Alibaba Cloud Intelligent Group CTO Zhou Jingren at the Zhongguancun Forum

In 2017, the birth of the Transformer architecture marked the beginning of the development of large models. Nowadays, large model technology has entered a stage of explosive development and is changing the entire application development chain. Zhou Jingren pointed out that the concept of Modelasaservice is being widely accepted, and the model will serve as an important production element for business and development system development.

Last November, Alibaba Cloud took the lead in proposing the concept of "Modelasaservice" in the industry and launched the open-source AI community's ModelScope, further lowering the threshold for AI development. Zhou Jingren said, "Modelasaservice is a new paradigm for the development of AI, and we hope that any developer, even elementary school students, can quickly experience the ability of artificial intelligence." In just six months, the total number of magic building community models has exceeded 900, with a total user volume of over 1.5 million, and a total model download volume of over 20 million.

At the same time, Alibaba is also actively investing in the research and application of large models. In 2018, Alibaba began developing large models and successively launched large models such as Tongyi AliceMind and Tongyi M6. In April of this year, Alibaba officially launched the Tongyi Qianwen Big Model and launched the Qianwen Partner Program. It is reported that over 200000 enterprises have applied to test the Tongyi Qianwen model, including Kunlun Digital Intelligence, Langxin Technology, Qianfang Technology, CICC Wealth, Shiji Group, Yonyou Network, and Asia Information Technology, all of which are building exclusive large models based on Tongyi Qianwen.

Zhou Jingren also emphasized that the development of large models is not just an algorithmic issue, but cloud computing has laid a solid foundation for the research and application of large models. He stated that Alibaba Cloud has formed a three-layer architecture of Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Model as a Service (MaaS), which can provide high-performance, low-cost computing resources and machine learning services for enterprises to train and reason large models. For example, Alibaba Cloud's high-speed and low latency network eRDMA can reduce latency to a subtle level, maximizing the computing power of AI clusters, Assist enterprises in efficiently training large models.

The high integration of cloud and AI is accelerating the arrival of the era of comprehensive intelligence. We hope to use the cloud platform and Tongyi Qianwen to help Qianxing Baiye create its own exclusive model around the concept of Modelasaservice.

Disclaimer: The content of this article is sourced from the internet. The copyright of the text, images, and other materials belongs to the original author. The platform reprints the materials for the purpose of conveying more information. The content of the article is for reference and learning only, and should not be used for commercial purposes. If it infringes on your legitimate rights and interests, please contact us promptly and we will handle it as soon as possible! We respect copyright and are committed to protecting it. Thank you for sharing.(Email:[email protected])