(Yicai Global) June 20 -- While other tech giants have been releasing their own ChatGPT-like artificial intelligence models, Tencent has announced a different approach by collaborating with its partners to develop targeted industry models.
At a press conference yesterday, Tencent's Cloud and Smart Industries Group unveiled the Industry Model Selection Store, which is built on the Tencent Cloud TI platform and provides customers with a one-stop Model-as-a-Service for model development and application.
The service is aimed at 10 major industries, including finance, culture and tourism, media and education, and currently offers more than 50 solutions.
Tencent CSIG Chief Executive Officer Dowson Tong said at a press conference that ecological co-construction is an effective path for AI development. Tencent aims to customize an exclusive large model for clients based on their actual needs, Tong said, as this can more effectively apply large model capability to the specific scenarios of particular industries and overcoming the obstacles of cost, data, and security that usually encountered in AI model applications.
Wu Yunsheng, vice president of Tencent Cloud and head of Tencent Cloud Intelligence, said the company’s large industry model is still in its early stages and requires cooperation with partners in many aspects to help gain a deeper understanding the know-how of specific industries and the needs of the specific application scenarios.
Tencent's basic object in mind is to help our customers solve problems with the most effective and cost-efficient ways, added Wu.
Tencent Cloud Vice President Wu Yongjian said that over the past decade, the company’s AI technology has been developing and the emergence of the ChatGPT-like artificial intelligence models have not make any fundamental changes in its technology development strategy.
From the demonstration cases and information at the press conference, Yicai Global saw that Tencent's industry AI model currently is mainly served for existing clients to upgrades their old applications, there are a handful of new clients indeed, but a significant increase in clients still need a qualitative improvement in both technology and applications.
Given that the firm’s rivals in China have not yet achieved a breakthrough in the commercialization of large models despite their intensive marketing, Tencent appears to have decided on taking a more cautious approach by working with its partners on specific industry models, probably a more winnable alternative approach.
Editor: Tom Litting