Future Me: Promoting a future-oriented outlook through conversation with our future self.

Large language models (LLM) transform the development of conversational agents for therapeutic purposes yet the benefits for mental health interventions have not fully been explored. Therefore, this study investigates the usage and acceptability of Future Me, a prototype of a chatbot harnessing generative AI, to facilitate future-oriented thinking. As part of an exploratory deployment study 20 participants engaged in short self-directed conversations with Future Me about current life obstacles. Following these conversations, semi-structured interviews were conducted with participants to assess the overall user experience, and acceptability, and to understand how participants used Future Me and how Future Me facilitated prospection. A thematic analysis of the suggestions of the system as well as the requests and reflections of the user indicates that participants used Future Me to explore barriers related to their own capability, opportunity, and motivation: Future Me responded by facilitating users to contemplate and prepare for behaviour change and providing positive reinforcement. Users reported positive outcomes such as improved mental clarity, enhanced self-reflection, and an increased sense of companionship and support. Limitations identified included user hesitancy around artificial intelligence for therapeutic purposes. This work contributes to a better understanding of the benefits and limitations of chatbots harnessing LLMs for the improvement of the user's mental health.