Voice vs. Text: Unveiling the Future of AI Explanation

This research introduces an exciting new information-theoretic framework for understanding how we can best explain AI systems! By comparing voice and text explanations, this model offers a reproducible foundation for designing and benchmarking explainability systems across different modalities, ultimately boosting user comprehension and trust.

• •The study introduces an information-theoretic framework for analyzing voice vs. text explanations.
• •Text explanations excel in comprehension efficiency.
• •Voice explanations show better trust calibration, especially with analogy-based delivery.