YouTuber Hank Green admitted to an unhealthy dependence on ChatGPT when creating videos
YouTuber Hank Green (3.2 million subscribers, Complexly channel) apologized for excessive use of ChatGPT when preparing videos after viewers spotted wording originating from a chatbot in the script. He promised to scale back production and return to content he writes himself.
YouTuber and writer Hank Green, who hosts educational videos for 3.2 million subscribers on the Complexly channel and other channels, publicly apologized for his growing reliance on ChatGPT when preparing videos. The controversy erupted after viewers noticed the unusual phrase “I appreciate the pushback” in a video, raising suspicions that the script had been created with help from a chatbot and that the model's response to one of the prompts had also been accidentally inserted into it.
Green initially stated in a (later deleted) post on X that the video had been produced “under a lot of pressure” and that he had used ChatGPT “for research for this script”; he described the disputed sentence as a response to a guest on the show. He subsequently published a more detailed apology on Reddit, where he wrote that he was “devastated” and had “let a lot of people down”. He maintains that he used ChatGPT only to “find academic texts and other sources” on the topics and that the actual words and opinions in the videos remain his own, but also acknowledged that criticism for “diluting” himself was justified. He also stated that he is not fundamentally opposed to AI, but has reservations about the technology's impact on the climate and how quickly companies working with AI are consolidating economic power.
According to Green, this is about more than one video — he admitted to an unhealthy level of “dopamine” from interactions with language models, which he described as harmful to himself and those around him, and said that this habit had disconnected him from how people respond to AI. He therefore promised to reduce how frequently he publishes on his channels and return more to content he writes himself — he mentioned a recent meditative video in which “writing was everything” and hinted that he would also make more unscripted videos featuring his own unedited thoughts.
The source focuses exclusively on the personal response and apology from Green; broader context (e.g. viewer reactions, the impact on channel viewership) is missing from the source. Details can be found in the source article.
Why it matters
The case shows how easily AI-assisted content can give itself away (unusual wording in a script) and the reputational risk this poses for creators who build their relationship with their audience on authenticity. It is also a personal account of patterns of dependence on interaction with LLM, which may serve as a warning for everyday users of similar tools at work.
Two audiences, two different impacts
What this means
For individuals
The case illustrates a specific warning sign of excessive LLM use — a growing sense of needing to interact with the model (“dopamine from LLM”) and losing a clear sense of one's own work process, which may also offer a lesson for everyday users of AI tools at work.
For a business
AI text used in published content without being identifiable as such can damage audience trust, as shown by a case in which viewers themselves spotted wording originating from a chatbot — for content creators and media outlets, this poses a risk of losing credibility regardless of how the AI tool was actually used.
Risks and complianceCheck the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.