Why Do We Need More Explainable AI?

Abdelkader Rhouati
Towards AI
Published in
5 min readJan 7, 2024

--

In the era of AI, where new models continue to emerge every day in all particularly sensitive areas, such as health and education, controlling these models becomes a necessity. The main problem with AI models is that they are designed as black boxes, which makes them impossible to control. Explainability or Explainable AI are techniques, principles, and processes introduced to solve this problem and make it possible to conceive a transparent, explainable, interpretable, equitable, and verifiable model.

--

--

Data scientist & Ph.D. researcher on AI. My area of expertise is around Deep Learning, NLP, and XAI