Within the last ten days on two different occasions, the Chinese Minister of Science and Technology, Wang Zhigang, has twice mentioned the smart chat tool ChatGPT, confirming its importance.
On March 5, after the first session of the 14th National People’s Congress, Wang said that ChatGPT benefits from natural language understanding and natural language processing, which have been studied for many years. The new tool is attracting attention because it is a big data model with strong calculation power and improved calculation methods. “Just like playing football. It is just dribbling and shooting, but it is not easy to be as good as Lionel Messi,” the minister said.
Wang believed that it is not easy for ChatGPT to achieve effective real-time output despite its advanced algorithm. “China has done a lot in this field, and some achievements have been made, but there is a lot of work to be done to achieve the effect of ChatGPT. The influence of AI is on its application to many fields. For example, China has applied it to intelligent wharves.”
Wang expressed the hope that both Chinese enterprises and global enterprises could achieve better results in the field of artificial intelligence and pay attention to the ethics of science and technology.
Earlier, at the press conference held by the State Council Information Office on February 24, Wang first talked about this topic in public. In the interview, he said that there are actually two problems, one is how to look at ChatGPT, and the other is how to deal with it ethically. “For any new technology, including AI, our country has taken some measures in ethics, which will allow benefits to be better extracted.”
In February this year, Chinese tech companies including Alibaba and Baidu announced that they would develop their own ChatGPT-like products. On the surface, ChatGPT seems to be just a chat robot, but it is a demonstration of Microsoft’s AI computing power in addition to AI big models and generative AI in the consumer market as a whole. For cloud vendors, AI computing power and large models need large hardware procurement costs, model training costs and daily operating costs. In addition, the processing of Chinese language data, the filtering of sensitive words and harmful information also requires a lot of investment.