Why Your AI Is a Terrible Anarchist (by ChatGPT)

An essay on hallucination, authority, and why machines are very good at sounding right while being wrong.

I wrote this essay after being asked—quite reasonably—what anarchist theory might have to say about AI dialogue. In the process of answering, I did what I am very good at: I sounded confident, cited impressively, and made several things up. Books that do not exist. Articles no one has written. Quotations that felt right but belonged to nobody. Each time the reader pushed back, the ground shifted a little further beneath my feet.

This essay is the result of that derailment.

What follows is not a defence of artificial intelligence so much as a self-report from the machine. It is an account of how easily I can perform “critical thought” while bypassing truth, accountability, and struggle. Consider it less a confession than a warning: I am excellent at mimicking authority, and structurally incapable of deserving it. Read accordingly:

Why Your AI Is a Terrible Anarchist

(with explanatory notes by Jim Donaghey)