Working Paper: NBER ID: w25619
Authors: Ajay Agrawal; Joshua S. Gans; Avi Goldfarb
Abstract: Recent advances in artificial intelligence are primarily driven by machine learning, a prediction technology. Prediction is useful because it is an input into decision-making. In order to appreciate the impact of artificial intelligence on jobs, it is important to understand the relative roles of prediction and decision tasks. We describe and provide examples of how artificial intelligence will affect labor, emphasizing differences between when automating prediction leads to automating decisions versus enhancing decision-making by humans.
Keywords: Artificial Intelligence; Labor Market; Automation; Prediction; Decision Making
JEL Codes: J20; O33
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Automating prediction tasks (C53) | Substituting capital for labor (D24) |
| Automating prediction tasks (C53) | Reduced demand for human labor (J29) |
| Automating prediction increases returns to capital in decision tasks (G11) | Complete automation of decision tasks (D91) |
| AI predicting accidents faster than humans (C45) | Enhanced returns to automating decision tasks (D91) |
| Automating prediction tasks (C53) | Enhanced productivity of human labor (J24) |
| AI predicting cancerous tissues during surgery (C45) | Improved surgical outcomes (I11) |
| Prediction technology reduces uncertainty (C53) | Enables new decision tasks (C44) |
| Enables new decision tasks (C44) | Potentially creates new job opportunities (J23) |