ChatGPT advice may undermine well-being – large cohort study highlights loneliness

The use of generative AI chatbots, such as ChatGPT, does not appear to be related to well-being in the same way, regardless of the purpose for which they are used. A recent study examined how different uses – chatting, asking questions, seeking advice, generating sentences, programming, organizing information, and translating – are connected to users' well-being and feelings of loneliness.

For six months, 3,533 Japanese adults were monitored. They reported both their usage patterns of ChatGPT and their experienced well-being and loneliness. The data was analyzed using a so-called parallel process latent change model, which aims to simultaneously track changes in AI usage and well-being and loneliness over time.

The results suggest that particularly social uses – such as asking questions and seeking advice – were indirectly associated with poorer well-being through increased loneliness. In other words, such use was linked to people feeling lonelier than before, which in turn was associated with a lower experience of well-being.

The study does not provide a definitive answer as to whether AI usage leads to loneliness or whether lonely people are particularly drawn to such use, but it emphasizes that not all AI usage is equal in terms of well-being. Distinguishing between different uses becomes central when considering the impacts of generative AI on everyday life and mental health in society.

Source: Investigating the relationships of ChatGPT usage for various purposes with well-being: 6-month cohort study, AI & SOCIETY.