Hi! I’m Cédric Colas, AI researcher exploring open-ended learning and computational creativity.
I’m interested in the study of artificial open-ended skill discovery — how an artificial agent can continuously grow a repertoire of skills throughout its lifetime. This interest led me to focus on intrinsically motivated agents that set their own goals, also known as autotelic agents.
I believe that for agents to develop such open-ended capabilities, they must be embedded in rich socio-cultural environments — interacting with peers, engaging with humans, and participating in a shared process of cultural evolution. We explore this idea further in our perspective article.
I completed my PhD at the Flowers Lab under the supervision of Pierre-Yves Oudeyer and Olivier Sigaud. My thesis, Towards Vygotskian Autotelic Agents: Learning Skills with Goals, Language and Intrinsically Motivated Deep Reinforcement Learning, is available here (also defense video).
I’m currently working at MIT with Joshua Tenenbaum and Jacob Andreas, developing autotelic agents that learn from humans and others using program synthesis methods.
Beyond my research, I explore creative applications of algorithms through various side projects.
See some of these explorations here.
