Our lab has as its mission to uncover the theoretical principles and neurobiological substrates that mediate adaptive behavior. To this end, we combine computational modeling which has its roots in artificial intelligence with empirical research where we measure the brain at work in naturalistic settings. In this talk I will put forward the notion that artificial neural networks provide cognitive neuroscientists with an ideal framework for modeling cognitive processes. Our results show that artificial neural networks accurately predict neural response patterns as the brain is engaged in challenging cognitive tasks, providing a new window on the neural implementation of such tasks. I will close with some thoughts on recent developments in this emerging area of research.