Explainability and the Fourth AI Revolution