Work

Combining Human Expertise with Artificial Intelligence: Experimental Evidence from Radiology

Year: 2023

Type: report

Abstract: Full automation using Artificial Intelligence (AI) predictions may not be optimal if humans have information not available to the AI (contextual information). We study human-AI collaboration using an information experiment with professional radiologists. Results show that providing (i) AI predictions does not improve performance on average, whereas (ii) contextual information does. Radiologists do not realize the gains from AI assistance because of errors in belief updating – they underweight AI predictions and treat their own information and AI predictions as statistically independent. Unless these mistakes can be corrected, the optimal human-AI collaboration design delegates cases either to humans or to AI, but rarely to AI assisted humans.

Source: National Bureau of Economic Research

DOI: 10.3386/w31422

Authors: Nikhil Agarwal, Alex Moehring, Pranav Rajpurkar, Tobias Salz

Institutions: National Bureau of Economic Research, Harvard University, Sinai Hospital, Yale University, Cowles Foundation for Research in Economics, Massachusetts Institute of Technology

Language: English

Topic: Artificial Intelligence in Healthcare and Education

Subfield: Health Informatics

Field: Medicine

Domain: Health Sciences

Open Access status: green

Funder: Alfred P. Sloan Foundation

Award: G5757865988

OpenAlex ID: W4383744587

Cites: 94

Cited by: 162

Related to: 10