"I don't trust AI": the role of Explainability in Responsible AI