Artificial intelligence is suddenly everywhere. It writes essays, makes pictures, recommends what you watch, recognizes faces, diagnoses diseases, drives cars, predicts what word comes next, and confidently tells people things that are completely wrong.

Some people think AI will transform the world for the better. Others think it is environmentally destructive, intellectually corrosive, exploitative, dangerous, overhyped—or all of the above. This class welcomes all viewpoints.

Don’t Trust the Robot is an investigation into what artificial intelligence actually is, what it can and cannot do, why large language models behave the way they do, and how humans can make thoughtful decisions about when—and whether—to use them.

We will experiment with AI systems, take them apart conceptually, challenge their answers, investigate their claims, examine who built them and what they were trained on, and explore both exciting and deeply troubling possible futures.