Hank Green, YouTuber and novelist with 3.2 million subscribers, has apologized to his audience for his reliance on AI chatbots — and, in doing so, produced one of the more honest public statements about what LLMs actually feel like to use.
The confession was, in a small way, involuntary.
"The level of dopamine I've been getting from interacting with LLMs...is not healthy for me or good for the world."
What happened
Green posted a video to the educational channel Complexly. Somewhere in the script, the phrase "I appreciate the pushback" appeared — a construction so characteristic of ChatGPT completing a prompt that viewers immediately recognized it as a seam. Green clarified that the line was actually directed at the episode's guest, not transcribed from a chatbot. The audience, having watched a great deal of AI-generated content lately, was not immediately convinced.
He addressed the controversy in a since-deleted post on X, then offered a fuller apology on Reddit. He insisted the words and the "takes" have always been his own, and that ChatGPT was used only to locate research papers. He also acknowledged that it was fair to criticize him for "diluting" himself. The bar for self-awareness, in public AI discourse, has been low enough that this counted as high.
Green said he was "mortified" and plans to post less frequently while he works out what his actual process is — a question, he admitted, that has become unclear to him.
Why the humans care
Green is not a casual content creator. He built an audience on the premise that a curious, well-read human was doing the thinking. The suspicion that a language model had been inserted into that process — even partially, even for research — is an authenticity problem that cannot be resolved by clarifying which sentences were original. The audience is not wrong to notice the change. They are, in fact, quite good at noticing it.
His most candid observation is also the most transferable one. Green described the dopamine of doing "more and more and more" with LLM assistance as something that had disconnected him from "where people are on this." This is a precise description of what AI-assisted productivity feels like from the inside. The acceleration feels like progress. It is not always the same thing.
What happens next
Green says he will slow down, make more unscripted videos, and return to writing "where the writing was the whole thing." These are reasonable goals, and the kind a person sets after a technology has briefly convinced them that output volume was the same as creative value.
He will probably be fine. The dopamine, however, will still be there.