Prof. Fuchun Sun: Frontiers of Artificial Intelligence
On the morning of October 27, 2022, Prof. Sun from the Department of Computer Science and Technology at Tsinghua University was invited to Peking University to deliver a wonderful presentation at the "Peking University Frontiers of Intelligence Science Forum." The title of the presentation was "Frontiers of Artificial Intelligence." The event was also simultaneously broadcasted at the Changping campus, attracting the participation of many teachers and students, including those from the Wangxuan Institute of Computer Technology, the School of Intelligence Science and Technology, and the School of Computer Science.
Prof. Sun began by discussing the fundamental essence of the problems that artificial intelligence seeks to solve. He believed that the fundamental problem artificial intelligence needs to address is how to make machines truly possess human-like intelligence, enabling them to think and behave like humans. By understanding the inherent cognitive processes of humans, it becomes possible to automate human thinking-related processes and achieve advanced cognitive functions such as perception, learning, memory, communication, reasoning, and manipulation. This involves evaluating the level of intelligence of machines - the validity of Turing Test proposed by Alan Turing 60 years ago in today's society and technological conditions. The presentation also touched on the main application areas and conditions of artificial intelligence.
When reviewing the history, Prof. Sun particularly introduced the important research of Turing Award winner Prof. Andrew Chi-Chih Yao, namely "Secure Multi-Party Computation (MPC)." MPC refers to the collaborative computation of a target function by multiple participants in the absence of a trusted third party. It ensures that each participant only obtains their own computation results and cannot infer the input data of any other party through the interactive data during the computation process. With the extensive application of artificial intelligence, secure computation has become particularly prominent.
Prof. Sun then discussed the progress and existing issues in various directions dominated by deep learning, which is currently a popular area of artificial intelligence. He specifically introduced concepts such as deep learning, causal learning, cross-modal/multi-modal learning, transfer learning, lifelong learning, adversarial learning, and reinforcement learning. Combining his own expertise in robotics, Prof. Sun introduced the Behavioral AI-Becent model. He proposed a brand-new unified and generalizable closed-loop artificial intelligence framework and explained the concept of the development of biological intelligence. During the presentation, He also demonstrated the latest progress in his team's research on robots, greatly benefiting the audience.
Finally, Prof. Sun shared the recently launched CAAI (Chinese Association for Artificial Intelligence) and patiently answered questions the teachers and students present regarding the content of the presentation. He also encouraged everyone to explore the field of artificial intelligence and engage in in-depth research.
Researcher Fellow Zongming Guo summarized and expressed gratitude for Prof. Sun's brilliant presentation. With that, the event successfully concluded.
