Editorial: From Explainable Artificial Intelligence (xAI) to Understandable Artificial Intelligence (uAI)

Abbass, Hussein ORCID: https://orcid.org/0000-0002-8837-0748, Crockett, Keeley ORCID: https://orcid.org/0000-0003-1941-6201, Garibaldi, Jonathan ORCID: https://orcid.org/0000-0002-9690-7074, Gegov, Alexander ORCID: https://orcid.org/0000-0002-6166-296X, Kaymak, Uzay ORCID: https://orcid.org/0000-0002-4500-9098 and Sousa, Joao Miguel C ORCID: https://orcid.org/0000-0002-8030-4746 (2024) Editorial: From Explainable Artificial Intelligence (xAI) to Understandable Artificial Intelligence (uAI). IEEE Transactions on Artificial Intelligence, 5 (9). pp. 4310-4314.

Preview

Accepted Version Available under License In Copyright. Download (429kB) | Preview

Abstract

In this editorial, we argue that the artificial intelligence (AI) community needs to escape the trap of explainable artificial intelligence (xAI) by growing more research on understandable artificial intelligence (uAI). We provocatively term xAI a trap because it has caused some AI researchers to see it as the “end” rather than a “means.” We will discuss why uAI is a better way forward and present a framework for uAI to define research directions that go beyond xAI. Let us first share where this concept emerged before introducing it.