When combinations of humans and AI are useful: A systematic review and meta-analysis

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Vaccaro, Michelle, Almaatouq, Abdullah, Malone, Thomas
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916458699685888
author Vaccaro, Michelle
Almaatouq, Abdullah
Malone, Thomas
author_facet Vaccaro, Michelle
Almaatouq, Abdullah
Malone, Thomas
contents Inspired by the increasing use of AI to augment humans, researchers have studied human-AI systems involving different tasks, systems, and populations. Despite such a large body of work, we lack a broad conceptual understanding of when combinations of humans and AI are better than either alone. Here, we addressed this question by conducting a meta-analysis of over 100 recent experimental studies reporting over 300 effect sizes. First, we found that, on average, human-AI combinations performed significantly worse than the best of humans or AI alone. Second, we found performance losses in tasks that involved making decisions and significantly greater gains in tasks that involved creating content. Finally, when humans outperformed AI alone, we found performance gains in the combination, but when the AI outperformed humans alone we found losses. These findings highlight the heterogeneity of the effects of human-AI collaboration and point to promising avenues for improving human-AI systems.
format Preprint
id arxiv_https___arxiv_org_abs_2405_06087
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle When combinations of humans and AI are useful: A systematic review and meta-analysis
Vaccaro, Michelle
Almaatouq, Abdullah
Malone, Thomas
Human-Computer Interaction
Artificial Intelligence
Computers and Society
Inspired by the increasing use of AI to augment humans, researchers have studied human-AI systems involving different tasks, systems, and populations. Despite such a large body of work, we lack a broad conceptual understanding of when combinations of humans and AI are better than either alone. Here, we addressed this question by conducting a meta-analysis of over 100 recent experimental studies reporting over 300 effect sizes. First, we found that, on average, human-AI combinations performed significantly worse than the best of humans or AI alone. Second, we found performance losses in tasks that involved making decisions and significantly greater gains in tasks that involved creating content. Finally, when humans outperformed AI alone, we found performance gains in the combination, but when the AI outperformed humans alone we found losses. These findings highlight the heterogeneity of the effects of human-AI collaboration and point to promising avenues for improving human-AI systems.
title When combinations of humans and AI are useful: A systematic review and meta-analysis
topic Human-Computer Interaction
Artificial Intelligence
Computers and Society
url https://arxiv.org/abs/2405.06087