Explainable AI: Learning from the Learners

TL;DR

This paper discusses how combining explainable AI with causal reasoning enhances understanding, trust, and robustness in high-stakes scientific and engineering applications by enabling learning from AI systems.

Contribution

It introduces a framework integrating XAI and causal reasoning to improve discovery, optimization, and certification processes in AI-driven scientific research.

Findings

Enables extraction of causal mechanisms from foundation models

Guides robust design and control in engineering tasks

Supports trust and accountability in high-stakes AI applications

Abstract

Artificial intelligence now outperforms humans in several scientific and engineering tasks, yet its internal representations often remain opaque. In this Perspective, we argue that explainable artificial intelligence (XAI), combined with causal reasoning, enables {\it learning from the learners}. Focusing on discovery, optimization and certification, we show how the combination of foundation models and explainability methods allows the extraction of causal mechanisms, guides robust design and control, and supports trust and accountability in high-stakes applications. We discuss challenges in faithfulness, generalization and usability of explanations, and propose XAI as a unifying framework for human-AI collaboration in science and engineering.