AI and Mental Health: Language Models Can Simulate Human Emotional States
Artificial intelligence language models can simulate human emotional states, opening up new possibilities for studying mental health, according to a new study.
Mental health disorders are becoming increasingly common worldwide. They are estimated to affect approximately 1.2 billion people by 2050. Against this backdrop, the scientific community is seeking new approaches to understanding the nature of mental conditions, their prevention, and the development of treatments.
Limitations of Traditional Methods
Researchers note that developing talk therapies remains challenging. Neither human clinical trials nor animal experiments can fully capture the complexity of mental processes, creating practical and ethical limitations.
Research from the Technical University of Dresden
A research team from the Technical University of Dresden (Germany) explored the use of large-scale language models (LLMs) to model cognitive and affective processes related to mental health.
Dr. Magdalena Weckenborg, head of the PsychoDigital Research group, noted:
"Our results demonstrate that large-scale language models are capable of reproducing patterns of human affective and cognitive processes under controlled conditions."
She added that such models can be used as a tool for a deeper understanding of psychological mechanisms and the discovery of new approaches, including the development of conversational psychotherapy.
