Session 3 – 27 May

AI in the brain

15:30 – 16:10

Modern AI is mostly based on so-called “deep learning,” which is a modern iteration of artificial neural networks: bigger, better and faster. Artificial neural networks in turn are strongly inspired by the networks of neurons that make up our brains. So does this mean we now understand how our brain works? And if not (because we don’t), how are artificial neural networks different from real neuronal networks? In this talk, Sander Bohté will discuss how artificial neurons differ from real neurons and how we can increasingly also compute with more biologically detailed neurons – with real implications for the proliferation of AI applications. Similarly, while deep learning is very successful in training big neural networks, the learning approach is considered biologically implausible. We’re trying to understand how real brains learn equally effectively as deep learning, resolving its ‘implausible’ parts.

Sander Bohté is a senior researcher in the CWI Machine Learning group and also a part-time professor at the University of Amsterdam and the Rijksuniversiteit Groningen. His research bridges the field of neuroscience with applications thereof in artificial intelligence as advanced neural networks. His work has been pioneering in the development of advanced and efficient spiking neural networks, and recent work has also developed biologically plausible deep learning and deep reinforcement learning models for neural network cognition. Together, his work is illuminating how the brain achieves the feats that deep learning does for AI.

Sander Bohté

Senior researcher at CWI