Photo by Yang Pudong
"Thank you very much to the World Artificial Intelligence Conference, Shanghai Association for Science and Technology, and Shanghai Science and Technology Development Foundation for awarding me the 2023 World Artificial Intelligence Conference Youth Excellent Paper Award." Zhang Yunwei, associate professor of the School of Physics of Sun Yat-sen University, said, "This award made me deeply feel everyone's attention to artificial intelligence, and also strengthened my academic path of continuing to engage in 'AI for Science' in the future." ”
At the 2023 World Artificial Intelligence Conference held in July, Zhang Yunwei's paper on "Machine Learning Combined with Impedance Spectroscopy to Predict the Aging of Lithium Batteries" won the highest award of the conference - SAIL Award. On August 24, at the 2023 World Artificial Intelligence Conference Youth Excellent Paper Award Ceremony and Young Scientist Forum held at the Shanghai Science Hall, she shared 10 excellent paper awards and 10 excellent paper nomination awards with 19 other young scholars.
From the perspective of the award-winning achievements, Chinese young scholars have shown vigorous vitality in the field of artificial intelligence research, not only the pursuit of cutting-edge hot spots, but also the ingenuity in the application landing, and more cross-integration will promote artificial intelligence to better empower technology.
"A lot of people asked me if this award-winning paper used a large model of the current fire." Zhang Yunwei responded in her speech, "I can tell you that there is no large model applied here, but sufficient and fast data processing is achieved through small models. This major is a young scholar in the direction of condensed matter physics, and has always focused his research on the field of lithium batteries. In her view, from smart phones to new energy vehicles, lithium batteries have entered human society on a large scale. However, these lithium batteries will face the fate of scrapping after 10-15 years of "service". Zhang Yunwei hopes that through new technology, the lithium battery during the "service" period will be accurately managed and the working life will be extended; Rapid detection of "decommissioned" lithium batteries to speed up recycling.
When Zhang Yunwei combined impedance spectroscopy technology with artificial intelligence, he ignited the spark of innovation: "The traditional way to detect batteries is to measure current and voltage, which is a macroscopic detection method. Electrochemical impedance spectroscopy, which is between microscopic and macroscopic, is a mesoscopic detection method that can know what is happening inside the battery and can more accurately feedback the aging process inside the battery. When we add artificial intelligence algorithms to impedance spectrum detection, on the one hand, we can shorten the detection time from the past few days to 15 minutes, and on the other hand, achieve more than 10 times the precision measurement accuracy higher than the industrial level. ”
"We very much hope to integrate more artificial intelligence into the lithium battery new energy industry." Zhang Yunwei is full of expectations for artificial intelligence to empower science and industry, "In the application process, we also need to overcome problems such as data collection and data security, which requires more young scholars to work together to continuously develop more optimized artificial intelligence models." ”
"Any task that can be serialized has the potential to be done by a large model, so it's a high ceiling. But the current big language model is still not perfect, which gives us the motivation to continue to move forward. "Ding Ning, Ph.D. of Tsinghua University, winner of the Youth Outstanding Paper Award of the 2023 World Artificial Intelligence Conference, is in the field of natural language processing, which is closely related to large models such as ChatGPT out of the circle. In his opinion, even the strongest big language models today are not perfect, "for example, it will involuntarily say a sentence, but the language model itself cannot explain 'why'." In each speech, the big model has no way to fact-check, which requires someone to judge whether the sentence is the truth. And the fact takes different forms on different occasions. ”
Ding Ning predicts that although the big language model needs to be continuously improved, it will still be the mainstream method of artificial intelligence in the next three years, and "we need to see its upper limit and study where the next breakthrough is." In this process, China's young scholars need to continue to innovate and contribute to the birth of new technologies, "In fact, in the most cutting-edge large model, there are many technologies from the field of Chinese intelligence as the basis, and there have been many large language models and related applications in China." But the core challenge remains the next breakthrough in the future. I hope that I can better understand the mechanism and properties of large models through theoretical research, make more powerful models, better align human values, and complete more complex scene tasks that have not been defined before."
"The Youth Excellent Paper Award of the World Artificial Intelligence Conference has lasted for four years, and 40 excellent paper award winners and 39 excellent paper nomination award winners have been selected. I believe that these 79 young scholars will bring great vitality and unlimited future to the development of Chinese intelligence and even the world's artificial intelligence. At the event jointly organized by the Municipal Association for Science and Technology and Shanghai Donghao Lansheng Exhibition (Group) Co., Ltd., Jiang Changjun, academician of the Chinese Academy of Engineering and chair professor of Tongji University, reviewed the process of the selection of the Youth Excellent Paper Award.
Over the past four years, the influence of the Youth Paper Award of the World Artificial Intelligence Conference has been increasing. This year, 235 papers were submitted at home and abroad, an increase of 80 over the previous session, demonstrating the enthusiasm of young scientists and scholars to participate. The number of submissions from famous overseas universities such as the United States, Singapore, Australia, and the United Kingdom accounted for 14% of the total number of submissions, demonstrating the international influence of the award. Many papers have been published in CVPR, IEEE series journals and other international artificial intelligence top academic conferences or authoritative journals, representing the highest academic level of young scientists and scholars in recent years, and many papers are already highly cited academic articles.
Jiang Changjun is full of expectations for young scholars: "I hope that these scientific seeds can be hatched in Shanghai and grow into towering trees for future industries." Han Zhiqiang, secretary of the party group and vice chairman of the Municipal Association for Science and Technology, and Zhang Ying, deputy director of the Municipal Economic and Information Commission, attended the event.