Summary of Integrating Expert Judgment and Algorithmic Decision Making: An Indistinguishability Framework, by Rohan Alur et al.
Integrating Expert Judgment and Algorithmic Decision Making: An Indistinguishability Framework
by Rohan Alur, Loren Laine, Darrick K. Li, Dennis Shung, Manish Raghavan, Devavrat Shah
First submitted to arxiv on: 11 Oct 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computers and Society (cs.CY); Human-Computer Interaction (cs.HC); Machine Learning (stat.ML)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed framework for human-AI collaboration in prediction and decision tasks leverages human judgment to distinguish inputs indistinguishable from feasible algorithms. By framing the problem as one where experts make judgments using side information not encoded in training data, the approach clarifies the challenges of human-AI collaboration. The algorithmic indistinguishability concept serves as a test for assessing whether experts incorporate such information and provides a method for selectively incorporating human feedback into predictions. The framework is shown to improve the performance of any feasible algorithmic predictor, with a case study in emergency room triage decisions demonstrating that physician judgments provide signal not replicable by predictive algorithms. This insight yields natural decision rules leveraging human expert and predictive algorithm strengths. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary AI can work better with humans when it’s hard for machines to tell things apart. Experts make judgments using information not programmed into machines, which is a problem for AI. A new way of working together lets humans help machines make predictions. This helps the machine make better decisions and uses the strengths of both humans and machines. |
