File Download
There are no files associated with this item.
Supplementary
-
Citations:
- Appears in Collections:
Conference Paper: AI Persuasion, Bayesian Attribution, and Career Concerns of Doctors
| Title | AI Persuasion, Bayesian Attribution, and Career Concerns of Doctors |
|---|---|
| Authors | |
| Issue Date | 6-Jan-2024 |
| Abstract | A substantial challenge to AI adoption is that humans often resist following AI. We study this AI aversion problem in the context of disease diagnosis and provide a new perspective, highlighting that disagreement between doctors and AI can result from multiple sources. When a doctor attributes her disagreement with AI to different sources, her willingness to follow AI varies. This feature allows us to manage doctor incentives by designing AI’s interpretability. In particular, making AI uninterpretable can actually enhance AI persuasion. When the doctor resists following AI due to career concerns, uninterpretability can decrease AI aversion and improve diagnostic accuracy. |
| Persistent Identifier | http://hdl.handle.net/10722/348076 |
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.author | Li, Hanzhe | - |
| dc.contributor.author | Li, Jin | - |
| dc.contributor.author | Luo, Ye | - |
| dc.contributor.author | Zhang, Xiaowei | - |
| dc.date.accessioned | 2024-10-04T00:31:18Z | - |
| dc.date.available | 2024-10-04T00:31:18Z | - |
| dc.date.issued | 2024-01-06 | - |
| dc.identifier.uri | http://hdl.handle.net/10722/348076 | - |
| dc.description.abstract | <p>A substantial challenge to AI adoption is that humans often resist following AI. We study this AI aversion problem in the context of disease diagnosis and provide a new perspective, highlighting that disagreement between doctors and AI can result from multiple sources. When a doctor attributes her disagreement with AI to different sources, her willingness to follow AI varies. This feature allows us to manage doctor incentives by designing AI’s interpretability. In particular, making AI uninterpretable can actually enhance AI persuasion. When the doctor resists following AI due to career concerns, uninterpretability can decrease AI aversion and improve diagnostic accuracy. <br></p> | - |
| dc.language | eng | - |
| dc.relation.ispartof | 14th POMS-HK International Conference (05/01/2024-06/01/2024, Hong Kong) | - |
| dc.title | AI Persuasion, Bayesian Attribution, and Career Concerns of Doctors | - |
| dc.type | Conference_Paper | - |