Shulou(Shulou.com)11/24 Report--

On May 20, at the closing ceremony of the 7th World Intelligence Congress held by Tianjin Haihe Media Center, host Jiang Changjian asked the big model. In just a few seconds, the continuation text generated by "Xunfei Spark" will "jump on paper."

In the face of the text content given by Xinfeld Spark Cognitive Model, the original author Liang Xiaosheng gave affirmation after careful evaluation: the content is concise, the text has a certain temperature, and it is almost "impeccable" from the emotion and thought conveyed.

In the face of human intelligence with thinking, reasoning, planning and creation, where will the development of a new round of cognitive large models take human beings?

At the closing ceremony of the 7th World Intelligence Conference in 2023, Jiang Changjian, professor of Fudan University, served as the host. Liang Xiaosheng, a famous contemporary writer and representative of "educated youth literature," Liu Cong, vice president of Xunfei University of Science and Technology and president of the Research Institute, Zhang Linghan, professor and doctoral supervisor of the Institute of Data Rule of Law of China University of Political Science and Law, gathered together. They not only investigated the text generation ability of Xunfei Spark Cognitive Model on the spot, but also from three perspectives of art, technology and academia. To jointly discuss the future development direction of general artificial intelligence.

"Xunfei Spark" continued to write "The World" and received a lot of praise

"I really think the sequel is good. "After the big model of Xunfei Star imitated the style of" The World "and continued to write a paragraph of text, the original author Liang Xiaosheng gave his evaluation.

Jiang Changjian (left) and Liang Xiaosheng (right) discuss the cognitive model of Xunfei Spark

"Because judging whether a paragraph of writing is good or bad, as Mr. Lu Xun said, first of all, from the rhetorical point of view, whether the words are concise, we see that it is concise, and it is not very possible to delete a few words from this kind of writing, saying that these words are completely redundant. "

When the big model meets literature and art, it makes amazing sparks. Behind this is the deep accumulation of Xinfeld Spark cognitive model in text generation and other abilities.

Liu Cong introduced the development of cognitive intelligence large model on the spot. Over the past few years, cognitive large models have continued to develop, but it was only after OpenAI released ChatGPT on November 30 last year that "intelligent emergence" was truly realized. "On May 6, Xunfei released the Spark Cognitive Model, which is also a general model that hopes that China will have its own and truly benchmark ChatGPT multi-task capability. "

Liu Cong introduces the cognitive model of Xunfei Spark on the spot

As the text-generating capabilities of large models evolve and iterate, new questions arise: How does AI-generated content define its copyright? Professor Zhang Linghan believes that a certain content has copyright because it is the intellectual achievement of human beings; however, machine-generated content has no originality, and generally does not think that copyright protection should be given to works generated by artificial intelligence, and hopes to better protect the original author in this way.

Zhang Linghan shares his views on the spot

Will big models steal jobs from humans in the future?

Faced with the amazing ability of cognitive intelligence large models to show multiple dimensions, host Jiang Changjian raised a question that many people care about: Will large models steal human jobs?

"Certainly not. Liu Cong replied. "Xunfei has always had the view that the future does not belong to AI, but to new human beings who have mastered AI. "

Liu Cong further pointed out that the "emergence of intelligence" brought by this cognitive intelligence model promoted the technological step of general artificial intelligence, but also brought about great subversion of production and life style, such as changing the mode of information distribution and acquisition, innovating the mode of content production, realizing all-natural interaction, realizing expert-level virtual assistant, becoming accelerator of scientific research work, subverting traditional manual programming mode, etc., which will have far-reaching social and industrial transformation impact.

"Are you worried that artificial intelligence is developing so rapidly? "Jiang Changjian threw the question to Liang Xiaosheng.

"I don't worry. "Liang Xiaosheng answered very simply.

Liang Xiaosheng said that literary works are combined through words and phrases of language. For example, the Chinese language has been used for more than 3000 years, but in recent years, new expressions have been created, such as "milk fierce," and netizens have created humor such as "laughing is not alive."

"Therefore, the creative power inspired by man's special life experience cannot be replaced by machines at present. "

Intelligent people will definitely run to artificial intelligence

"Will robots have the consciousness to actively modify? "Liang Xiaosheng asked.

Liu Cong said, At present Xinfeld spark cognitive model has been able to realize relevant functions, For example, in the equipped Xinfeld AI learning machine, Can like the teacher layer upon layer to comment on the child's composition, For example, the child a paragraph of composition description is not vivid enough, Can through the embellishment function to provide modification reference, In order to achieve targeted partial modification.

Faced with such rapid iterative evolution of AI, Jiang Changjian raised the question: Can artificial intelligence generate self-awareness? Should mankind be cautious about its development?

"Whether it's data or products, behind artificial intelligence are people, and I don't think machines will become self-aware in the short term. "Liu Cong said.

Zhang Linghan suggested that at a time when artificial intelligence has a great impact on social life, work, study and other fields, we should face it with an embracing attitude and believe that intelligent people will run to it.

In this World Intelligent Conference, in addition to Liu Cong participating in the closing round-table theme forum, Liu Qingfeng, Chairman of USTC Xinfeld, was invited to attend the opening ceremony and Innovation Development Summit of the Conference and delivered a keynote speech; Li Weibing, Deputy General Manager of Intelligent Automobile Business Department of USTC Xinfeld, attended the "International Summit Forum on Innovation and Development of Internet of Vehicles Pilot Application" and delivered a keynote speech.

At the intelligent science and technology exhibition of the 7th World Intelligent Conference held at the same time, the exhibition of core products and solutions in the fields of AI + city, AI + medical treatment, AI + industry, AI + office, etc. also showed the new exploration and new achievements in the fields of AI enabling society and people's livelihood.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.