The Thought Police: Should We Monitor AI’s Inner Dialogue?
193