HAI is a project with the goal of creating a dual-agent cooperation game to support the research on Human/AI collaboration dynamics.

The team mainly focuses on asymmetrical gameplay where two agents are given different interactions/information to achieve a common goal. The game is designed such that each agent can swap with an AI so that we provide an environment where researchers can easily compare Human/AI teams with Human/Human and AI/AI teams. Researchers are able to collect data with fun, engaging, and strategic gameplay. Embedded game metrics are also employed as the primary research data source.

Our clients are Timothy Verstynen and Michael Tarr from the Department of Psychology at CMU.

Our Deliverable


Players’ Tasks