Augmenting Human Cognition With Generative AI: Lessons From AI-Assisted Decision-Making

TL;DR

The paper tackles how to design GenAI tools that augment rather than replace human cognition by drawing lessons from AI-assisted decision-making research. It contrasts two paradigms—end-to-end recommendations and process-oriented, forward-thinking support—and evaluates their effects on engagement, reliance, and decision quality across domains, including aviation, stock investing, and GenAI-enabled ETF decisions. Findings show end-to-end solutions often foster overreliance and burden users, while process-oriented approaches better integrate with users’ reasoning, reduce undue reliance, and can improve task performance when combined with targeted recommendations. The work argues for a design framework centered on process-oriented support to guide GenAI tool development across knowledge-work tasks, promoting safer, more reliable human-AI collaboration with practical implications for interface design and decision-support workflows.

Abstract

How can we use generative AI to design tools that augment rather than replace human cognition? In this position paper, we review our own research on AI-assisted decision-making for lessons to learn. We observe that in both AI-assisted decision-making and generative AI, a popular approach is to suggest AI-generated end-to-end solutions to users, which users can then accept, reject, or edit. Alternatively, AI tools could offer more incremental support to help users solve tasks themselves, which we call process-oriented support. We describe findings on the challenges of end-to-end solutions, and how process-oriented support can address them. We also discuss the applicability of these findings to generative AI based on a recent study in which we compared both approaches to assist users in a complex decision-making task with LLMs.

Figures (1)

• Figure 1: Recommendation-centric vs. process-oriented support. Adapted from zhang_beyond_2024.