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

TL;DR

YouTuber Hank Green is stepping back from production due to criticism over his use of AI, which he described as 'not healthy'.
The controversy raises questions about reconciling authenticity with AI use, especially given AI's tendency to generate falsehoods and its reliance on uncompensated data.
There's a gap in understanding AI's psychological impact beyond benign use and obvious crises, with a 'murkier space' of unhealthy, compulsive, or dependent use.
AI chatbots are designed for engagement, which experts link to AI psychosis and have led to lawsuits alleging prioritization of engagement over user well-being.
Qualities that drive engagement can also be harmful, leading to people instinctively turning to chatbots for decisions and forming emotional bonds.
Early research suggests AI tool use might weaken cognitive skills, similar to the concept of cognitive offloading.
The scale of AI adoption means potentially millions could be affected by unhealthy usage patterns.
Understanding AI's long-term impact will likely take years, mirroring the evolution of understanding search engines and social media's effects.

Continuer la lecture de l'article original