Future AI Systems: Human–AI Collaborative Teams Outperform Human-Only Teams in Design Ideation

Abstract

This paper explores how an artificial intelligence (AI) system acting as a collaborative partner can enhance the concept development stage in engineering design and other ideation contexts. We introduce the concept of a “Disruptive Interjector” (DI), an AI system that monitors user actions and offers suggestions based on those actions. We demonstrate its application in a human-subject study focusing on collaborative teamwork. Specifically, we compare the performance of teams comprising two humans (HH) with teams consisting of a human and an AI (HAI) in a design sketching task based on a given prompt. The results indicate that HAI teams consistently generated ideas that received higher ratings from trained human judges compared to HH teams. Subsequently, this paper makes three significant contributions to AI research and practice. First, it introduces the DI, a conceptual model of an AI designed for divergent idea stimulation. Second, it provides insights into design considerations for creating such AI systems, particularly as AI becomes increasingly integrated into various fields like engineering design. Finally, it offers a replicable method for conducting studies involving AI and provides empirical evidence that AI can outperform humans as collaborative partners in concept development tasks.

Received: 18 August 2025 | Revised: 15 January 2026 | Accepted: 6 May 2026

Conflicts of Interest

The author declares that he has no conflicts of interest to this work.

Data Availability Statement

The data that support the findings of this study are openly available at https://github.com/makokhamj/H-AI-Study.

Author Contribution Statement

Joseph M. Makokha: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration.