Researchers suggest asking AI to disagree with workers improves outcomes
A study of knowledge workers at a U.S. public university found that intentionally using generative AI to challenge ideas can improve performance.
Study examines critical thinking risks
Scholars warn that relying on quick and confident answers from generative artificial intelligence can lead to low-quality solutions. This phenomenon is known as cognitive offloading because it allows users to surrender their own critical thinking skills.
Interviews reveal better results with friction
Researchers interviewed 45 employees who used these tools for generating ideas and solving complex problems. The team analyzed chat logs and discovered that workers who asked the AI to contradict their thoughts saw significant performance gains.
Previous loan evaluation research matches findings
Earlier work on human-AI collaboration for loan evaluations produced similar positive results when friction was introduced. The current project extends these insights to broader knowledge work tasks like designing novel solutions.
Reported by one outlet
Only one outlet has published this. Nothing here has been checked against a second report, so read it as that outlet's account and follow the link below for the original.
Reported by
1 independent outlet. Headline as published. Links open the original report.