Prof at Mila/UdeM. Maximally Beneficial AGI asap :). Continual learning, OoD generalization, neuro-AI, AI scaling, AI alignment. Opinions are my own.