Date: October 28, 2024 (Monday)
Speaker: Prof Carlos Montemayor , San Francisco State University
Chair: Dr Frank Hong, The University of Hong Kong
Abstract:
The two prevailing paradigms of artificial intelligence, based on symbolic systems and neural networks, need further refinements in order to qualify as general artificial intelligence. I will argue that the best way to interpret and improve these models is by defining general intelligence as the integration of solutions to problems generated by the cognitive needs of an autonomous agent. This allows us to address various limitations of current models, for instance, the importance of embodied cognition for general intelligence; the hierarchical structure of our plans and goals and the way they shape the solution to a variety of problems; the different domains of our cognition, epistemic, moral, and aesthetic; and the relation between our need for autonomy and human dignity. I will consider several options concerning the logical space this proposal occupies, emphasizing why general intelligence should entail cognitive freedom and autonomy.
