Experts generally agree that, before we might consider artificial intelligence (AI) to be truly intelligent —that is, on a level on par with human cognition— AI agents have to pass a number of tests.