My research aims at improving dialogue between humans and machines. That’s what people sometimes mean when they say “AI”, but the news is that to participate in dialogue is much harder than to play an Atari game. That is because dialogue is a form of joint action and it requires a great deal of coordination on all levels: from timing to meaning. I position my research in a field of computational linguistics in a broad sense: I build models that reproduce human behaviour, therefore trying to understand how our cognition works. And one of the by-products of this is, I believe, a better AI.