AI for Understanding Human Intelligence

Description
Human infants acquire various cognitive abilities in the first few years of life. Although the developmental dynamics of their behaviors have been closely analyzed, what neural, bodily, and social mechanisms guide the development remain a mystery.
In this course, I will introduce AI and robotics approaches to understanding the underlying mechanisms for infant development. The approach called cognitive developmental robotics aims to elucidate the principle of human intelligence by designing artificial systems that learn and develop like infants. In contrast to the analytical approach in neuroscience, cognitive science, and psychology, this constructive approach has the potential to uncover a unified principle of intelligence. 
The course consists of three parts: lecture, hands-on , and presentation. In the first week (Sessions 1-5), I will give lectures on how AI and robotics technologies can be used for investigating cognitive development. Computational studies using neural networks and humanoid robots will be introduced to explain how neural, bodily, and social mechanisms interact to guide cognitive development. In the second week (Sessions 6-17), students will work on hands-on projects  to learn practical challenges in pursuing the above studies. Students divided into groups will address one of the following topics: (a) programming of neural networks to test a computational theory of cognitive development, and (b) robot/VR experiments to examine neurodiversity in cognitive development. At the end of the second week (Sessions 18-20), students will give a presentation about their hands-on projects. Students will discuss how the theories of cognitive development can be tested using neural networks and/or a robot/VR and what they have achieved and learned from their projects.