PPP 518 | Why Better AI Prompts Aren't Enough for Better Decisions, with author Cheryl Strauss Einhorn

Summary In this episode, Andy welcomes Cheryl Strauss Einhorn, founder and CEO of the decision sciences company Decisive and author of The Human Edge: Smarter Decisions in the Age of AI. Cheryl has spent decades helping people make better decisions, and her message here is a simple one: AI can gather, summarize, and compare, but it can't know what matters most unless we do the thinking first. Andy and Cheryl talk about why "AI first" can quietly turn into AI only, and why a poor answer from AI is actually useful feedback about your own thinking. Cheryl explains why separating the what of a decision from the why can change the path forward, shares her vision of success question, and walks through how human bias and AI bias reinforce each other. You'll also hear a practical way to use AI personas to pressure test your thinking before a big stakeholder conversation, plus what parents can do to help kids strengthen their decision-making muscles instead of letting them atrophy. If you're looking for practical ways to make smarter decisions in the age of AI, this episode is for you! Sound Bites "AI is about patterns, but we're about purpose." "We need to be the chief deciders in our own life." "We need to make sure that we actually have a way to push back and check the veracity of what it's giving us because it's a known liar." "And so the authenticity of human interaction is what actually builds trust, strengthens relationships, and really gives us a lot of the connections that are important to us in our home life and our work life." "You don't need to have a perfect prompt for AI." "AI is going to give you other people's answers." "'Knowledge is power' isn't as true now. It's what you are able to do with the knowledge, and it's why you want to collect the knowledge in the first place." "I almost never would accept AI's first answer." "We tend to have evolving hypotheses unless we really build an audit trail of our thinking." "So in the world of medicine, a fever tells you something is wrong, but it tells you nothing about what is wrong or where to look." "So first we know AI's a sycophant, right?" "If it's 99% accurate in its conclusion, but its underlying data is the wrong data set, you've got the wrong answer." "I would say the more that I've studied these cognitive biases, these mental shortcuts, the more I realize that we see the world through a dirty windshield." "Our brains are muscles, and just like we exercise to strengthen our muscles, if we don't exercise, they atrophy." Chapters 00:00 Introduction 02:00 Start of Interview 02:12 Growing Up Surrounded by Questions 04:30 What Leaders Should Watch Out for in "AI First" 07:07 When AI Confidently Makes Things Up 09:04 How AI First Turns into AI Only 09:20 Why Polished Doesn't Mean Authentic 11:59 A Prompting Problem or a Problem Definition Problem? 14:37 The Vision of Success Question 15:45 Why a Bad Answer Is Useful Feedback 19:06 Separating the What from the Why 20:32 Research: More Information Isn't Better Judgment 24:35 Progressive Prompting and Never Accepting the First Answer 25:18 Fabricated Quotes and the Trouble with Attribution 26:50 Documenting Assumptions and Building an Audit Trail 29:25 Committing to Your Own Thinking 32:29 How Human Bias and AI Bias Reinforce Each Other 34:55 Knowing About Biases Doesn't Make You Immune 36:51 Using AI Personas to Challenge Your Thinking 40:20 AI and Stakeholder Management 42:42 Helping Kids Become Better Decision-Makers 44:25 End of Interview 44:51 Andy Comments After the Interview 48:13 Outtakes Learn More You can learn more about Cheryl and her work at AREAMethod.com. For more learning on this topic, check out: Episode 460 with Joe Sutherland. It's an interesting look at the intersection of AI, data, and decision-making, and a great follow-up to this discussion. Episode 381 with Jim Loehr. One of Andy's favorite conversations about decision-making, from a remarkable figure in the human performance world. Episode 99 with Mike Roberto. A conversation about hi