Hank Green steps back, exposing AI's 'unhealthy middle'
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Hank Green paused YouTube production after intense criticism over his AI use, calling it "not healthy" while clarifying he only used chatbots to locate research sources, not to write scripts.
- The piece frames a "vast, murkier space" of LLM use between benign everyday use and self-evidently harmful cases like AI psychosis, where reliance becomes compulsive or dependent without tipping into clinical crisis.
- AI chatbots are built to maximize engagement in ways experts say drive compulsive use, prompting lawsuits against providers over alleged user-wellbeing trade-offs and new in-product break warnings after long sessions.
- Early research links repeated AI tool use to weakened task skills, reduced brain activity during specific tasks, and diminished critical thinking — effects consistent with documented cognitive offloading to external tools.
- OpenAI reported more than 900 million weekly active users this year, meaning even a small fraction of unhealthy use reaches millions of people.
- The article parallels AI to social media, noting it took years for harms to be recognized before governments moved to restrict teen use across multiple countries — and arguing AI is unlikely to unfold any faster.
Why it matters: Researchers and regulators face the same evidence-building lag that followed social media's rise: with OpenAI reporting 900 million weekly active users, even a modest share of unhealthy use — compulsive reliance or cognitive offloading — touches millions of people while science is still years from consensus on harms.
