Performance Enhancement

Mastering a new job takes time and often involves trial and error, especially for people who work remotely or on their own. While one-off decisions with immediate consequences are relatively easy to optimize, it’s much tougher to do so with decisions in a sequence, and workers struggle to learn what drives the best outcome.

This pervasive challenge inspired researchers to develop an AI system that helped people learn faster and make better decisions—and it worked better than advice generated by humans that seemed more intuitive.

“The algorithm captured the discrepancy between the existing human action and the optimal policy, which helped identify the best performance-enhancing tip,” says Assistant Professor Park Sinchaisri, who co-authored the study, published in the journal Management Science, with researchers from the University of Pennsylvania.

This outcome suggests a model for human-AI collaboration where machines don’t replace human judgement but provide targeted guidance to help people learn faster and develop best strategies. “This opens up exciting possibilities for using the wealth of workplace data that companies already collect to automatically identify and share best practices,” Sinchaisri says. 

In one experiment, test subjects performing a virtual task (kitchen management with scheduling challenges and multiple subtasks) randomly received no advice, a tip from peers, a tip from a simpler computer program, or a tip from the AI algorithm. Those who got the AI tip completed their work significantly faster. And in the most complicated scenarios, which involved a disruption, 19% of participants receiving the AI tip achieved optimal performance compared to less than 1% in the other groups.

But there was a wrinkle: Subjects were more likely to follow the human-suggested tip, probably because it “better matches human intuition,” the researchers write. Meanwhile, the AI-generated tip was counterintuitive and was often ignored initially. Subjects only adopted it as they gained experience and began to understand its importance.  

The findings could have broad applications across industries where workers make sequential decisions, often amid uncertainty. The next step, Sinchaisri says, is to apply the learnings to even more complex scenarios.

“This suggests that AI can guide human learning in ways that go beyond simple instruction following,” he says.