Abstraction and Cognition with Prof. Yoshua Bengio

Join us in this enlightening talk with Prof. Yoshua Bengio as he delves into the heart of deep learning. Exploring the concept of multiple levels of representation and abstraction, Prof. Bengio illuminates how these layers shape our understanding of artificial intelligence. He bridges the gap between human cognition and machine intelligence, offering insights into how conscious perception can provide hints for higher levels of abstraction in AI. This talk further unravels how these high levels of representation are crucial in disentangling the complexities mixed up in raw data. Prof. Bengio emphasizes the importance of explanatory variables and causal mechanisms in achieving effective generalization, particularly in out-of-distribution translation and transfer. An essential watch for anyone interested in the future of AI and its intersection with human cognitive processes. #OXML23 #AI4GG #SDGs #MachineLearning #AI