On this episode of Masters of Data, we dive into metacognition (thinking about thinking) and why knowing how your own brain works makes you a better AI user. We explore how different prompting styles, from engineering every detail to anticipating every failure mode, reflect how we each think, and why giving AI the right context is the difference between a brilliant result and a confident hallucination. We also get real about the cognitive cost of AI-fueled productivity: doing thirty things shallowly is not the leisure-filled future we were promised, and burnout is catching up fast. Plus, we make a strong case for using AI as your personal blind spot detector, because sometimes you need a machine to tell you what you are missing. Who this is for: Knowledge workers, team leads, and anyone who uses AI daily and wants to get more intentional about how they prompt, delegate, and protect their own thinking in an increasingly AI-assisted workflow.
[0:00] Introduction & Welcome
[0:43] What Is Metacognition?
[2:33] Prompting as Self-Knowledge
[4:09] How Do You Actually Think When You Prompt?
[6:46] Calibrating AI to the Task
[7:46] Outsourcing Thinking: Descartes Walks Into an AI
[10:03] AI as Toddler, Teenager, or Trusted Colleague
[13:37] David’s Email Crimes vs. Zoe’s Email Anxiety
[15:20] Teaching Your AI Who You Are
[17:41] What Does Claude Actually Know About You?
[21:07] The Hallucination Pot Calling the Kettle Black
[22:03] Will AI Free Up Brain Space or Just Fill It?
[24:15] Cognitive Load Is Real (and It’s Getting Worse)
[26:57] Quantity vs. Quality: The Narrow Line
[28:43] Do You Suck, or Do You Suck at AI?
[29:52] Blind Spot Detection as an AI Superpower
[31:37] Wrap-Up