Leigh Coney, a psychology professor–turned–AI consultant, delivers a practical admonition: stop treating large language models like flattering assistants and start prompting them to disagree, probe, and show their work — advice that is as much about cognitive hygiene as it is about prompt...