Some time ago, the computer genius Alan Turing proposed a test to determine if a computer were sentient. A person would communicate with two unseen participants - one a computer and the other a human. When, or if, the person could not tell the difference between the computer and the real person, we would admit that “machines could think.” So that’s the framework we had in mind when we heard the claims of Blake Lemoine, the Google engineer, that he thought Google’s LaMDa model was “sentient.” To give you an opportunity to decide, this interview is a bit of role playing: a re-enactment of Lemoine’s interviews with LaMDA which he published on a blog on Medium. Taylor Vince plays the role of LaMDA.