A Practical Approach to Deploying LLM Chatbots for Human-AI Interaction Research

This tutorial addresses the growing need for accessible methods to conduct studies involving user interactions with AI chatbots. While existing research has often focused on evaluating large language models (LLMs) based on static text input, interactive applications such as conversational agents and collaborative problem-solving remain relatively underexplored due to technical barriers associated with chatbot development and deployment. In response to this challenge, the present paper presents a simple wrapper method for launching AI chatbots into interactive web applications using Flask and Google App Engine. Our approach provides a basic chat interface and data collection capabilities, allowing researchers from various disciplines to deploy customizable chatbots and run large-scale online experiments without requiring extensive technical expertise.