The University of Macau (UM) announced in a Tuesday statement that a joint study published in the international journal Trends in Cognitive Sciences concluded that artificial-intelligence (AI) is unlikely to provide a long-term solution to human loneliness.
According to the Wikipedia, Trends in Cognitive Sciences is a monthly peer-reviewed review journal published by Cell Press. It is one of 14 journals in the Trends series
The research was conducted by Christian Montag, associate director of UM’s Institute of Collaborative Innovation, Michiel Spapé, associate professor at UM’s Centre for Cognitive and Brain Sciences, and Benjamin Becker, professor in the Department of Psychology at the University of Hong Kong, according to UM.
Loneliness, affecting one in six people globally, has been recognised by the World Health Organisation (WHO) as a public health concern with serious physical and mental consequences, the statement said, while some studies suggest individuals are increasingly turning to AI companions or chatbots as substitutes for friends or therapy, the three researchers caution against viewing AI as a cure.
The statement quoted Montag as remarking that “if a person is lonely, they long for the physical presence of someone close to them. This is very hard to mimic with the available chatbots. Large language models operate mostly via text communication. Clearly, this is not sufficient to provide lonely people with the best cure for loneliness being direct support in person.”
The statement added that the study emphasises the importance of human touch and direct interpersonal connection, noting that while AI may support individuals in crisis when human help is unavailable, it cannot replace genuine human relationships.
The authors stress the need to make AI safer before deploying it in psychological support contexts, and highlight that investments in AI should not overshadow community building and human-to-human care, the statement noted.
This infographic provided by the University of Macau (UM) on Tuesday shows the human-human and human-large language model (LLM) interactions regarding loneliness.
