Should We Really Care About Explainability In AI?