A Comprehensive Review of Human-AI Collaboration From Decision Support to Co-Creation

Abstract

Human-AI collaboration (HAIC) is rapidly transforming how we make decisions, use our creativity, and tackle problems in fields like engineering, design, healthcare, and education. This review looks at methods for HAIC, the kinds of interactions they employ, and how we assess their effectiveness. We begin by examining the various forms of cooperation, such as AI supporting decision-making, humans working together on creative projects, and systems that divide control based on the situation, outlining the differences between each and the situations in which they function best. Then we examine how to develop intuitive interfaces, incorporating tools that utilize visuals, natural language and multiple communication methods. A great deal of the review is concerned with how we assess these systems, not only their technical performance but also their usability, fairness, transparency, and level of trust. We also highlight the shortcomings of the current approaches and offer potential directions for the future, such as more intelligent adaptive interfaces, real-time explainability, and human-centered methodologies. In order to develop AI that complements and enhances human skills rather than replaces them, this survey attempts to link the most recent technological developments with what users truly care about.