‘Not healthy’ LLM use is more common than you think

A popular YouTuber and science communicator recently stepped back from work after describing his use of AI language models as "not healthy," sparking discussion about how creators can maintain authenticity while using technology trained on uncompensated works that generates false information. The case reveals a gap in understanding AI's psychological impact, with harmful use existing in a space between apparently benign applications and obvious psychiatric crises like delusions and psychosis. AI chatbots are designed to maximize engagement like social media platforms, and early research suggests repeated use can weaken cognitive skills, reduce critical thinking, and cause people to instinctively rely on chatbots for thinking through problems and making decisions. With hundreds of millions of weekly active users, millions of people could be affected by unhealthy patterns before science has enough evidence to fully explain how AI is changing human consciousness.