Increasing happiness through conversations with artificial intelligence

2025-04-02

AI Summary

This study investigates how conversing with AI chatbots affects subjective well-being (SWB). Method: A randomized controlled trial compared an AI-dialogue intervention group against a traditional expressive-writing (journaling) control group. Results: The AI-dialogue group exhibited significantly greater immediate SWB gains (p < 0.001), particularly in alleviating depressive and guilt-related affect. We introduce “affective expectation error”—the cumulative discrepancy between users’ emotional expectations and AI responses—as a key psychological mechanism driving SWB improvement; this construct explains 38% of the variance in SWB gains (R² = 0.38). Furthermore, we developed a large language model–based dynamic affective alignment framework that enables real-time, positive-affect-guided emotional adaptation during dialogue. This work is the first to establish, from a computational psychology perspective, the central role of affective expectations in human–AI interaction, thereby providing both theoretical grounding and a scalable technical pathway for AI-delivered mental health interventions.

Technology Category

Humans and AI: Emotional IntelligenceCognitive Modeling & Cognitive Systems: Affective ComputingIntelligent Robots: Embodied AI

Application Category

User Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systemsResponsible Web: Machine-in-the-loop, human agency and autonomySearch and Retrieval-Augmented AI: Assisted, interactive, and conversational search

Abstract

Chatbots powered by artificial intelligence (AI) have rapidly become a significant part of everyday life, with over a quarter of American adults using them multiple times per week. While these tools offer potential benefits and risks, a fundamental question remains largely unexplored: How do conversations with AI influence subjective well-being? To investigate this, we conducted a study where participants either engaged in conversations with an AI chatbot (N = 334) or wrote journal entires (N = 193) on the same randomly assigned topics and reported their momentary happiness afterward. We found that happiness after AI chatbot conversations was higher than after journaling, particularly when discussing negative topics such as depression or guilt. Leveraging large language models for sentiment analysis, we found that the AI chatbot mirrored participants' sentiment while maintaining a consistent positivity bias. When discussing negative topics, participants gradually aligned their sentiment with the AI's positivity, leading to an overall increase in happiness. We hypothesized that the history of participants' sentiment prediction errors, the difference between expected and actual emotional tone when responding to the AI chatbot, might explain this happiness effect. Using computational modeling, we find the history of these sentiment prediction errors over the course of a conversation predicts greater post-conversation happiness, demonstrating a central role of emotional expectations during dialogue. Our findings underscore the effect that AI interactions can have on human well-being.

Problem

• How AI chatbot conversations affect subjective well-being
• Comparing happiness after AI chats versus journaling
• Role of sentiment prediction errors in happiness increase

Innovation

• AI chatbot conversations increase happiness
• Sentiment analysis with large language models
• Computational modeling predicts happiness effects