The First Time AI Blew My Mind

For the first dozen years we were parents, we lived in a 120-year-old house in Mountain View. With four rambunctious daughters, mornings are... special.

One morning, I was making breakfast—calmly, like a short-order cook who knows everybody’s egg preference—when I heard a BANG from down the hall. The kind of bang that makes you freeze.

I did what any responsible father would do: I pretended I didn't hear it. (Just keep flipping the eggs. Maybe it was nothing.)

"Dad?!"

It was not nothing.

I walked down the hall to find blood, broken glass, and a shattered window. Not just any window—one of the original Victorian windows that came with the house. Irreplaceable. My daughter stood there, terrified, hand cut, surrounded by the evidence.

One of the girls asked, "Aren’t you a professional problem solver?" Yes, but…

Here's the thing about being a professional problem solver: it doesn't work when your heart rate is through the roof and you're staring at your kid's blood on the floor. My brain defaulted to the only pattern it knew. How many days does she lose the iPad? A week? Two weeks? Ground her until she's thirty?

I could feel myself about to say something I'd regret.

The Reflexive Invitation

This is the part that still surprises me when I tell the story. Without really thinking about it, I pulled out my phone and opened ChatGPT.

Not because I had some brilliant strategy. Not because I remembered a framework from a keynote. I'd just... started doing this. Inviting AI into moments where I felt stuck. It had become reflexive.

I typed (okay, dictated—I was too frazzled to type):

"I just discovered my daughter broke an irreplaceable window in our house. There's broken glass everywhere, she cut her hand, emotions are running high. Act like a child developmental psychologist with a specialty in parenting. Give me 10 terrible ways to respond to this situation."

Why terrible ways? Because when you're activated—when your amygdala is running the show—you can't think straight enough to evaluate good ideas. But you can definitely spot bad ones. (This is a trick I'd picked up somewhere. Ask for the wrong answer, and your judgment kicks in. Why not have some fun with it?)

What Came Back

ChatGPT gave me exactly what I asked for. Ten gleefully terrible responses, including:

• Ground her indefinitely with no clear path to redemption

• Call 911 and report her to the authorities (I laughed out loud at this one)

• Make her pay for the window replacement out of her allowance for the next 47 years

• Yell until she cries, then feel guilty about it later

As I read the responses aloud, we all started chuckling a little. The tension eased. Even better, buried in the terrible suggestions was something I never would have thought of:

Have your daughter prepare and deliver a "homeschool lesson" as a teacher to her younger sisters on why household rules exist and why they matter for everyone's safety.

I stared at it.

This wasn't punishment. This was... ownership. Connection. Responsibility. Learning. She wouldn't be shamed into compliance—she'd be elevated into teaching. She'd have to actually think about why the rule mattered, articulate it to her sisters, and in doing so, internalize it herself.

I would never have come up with that. Not in that moment. Not with my heart pounding and everybody waiting for my "professional-grade solution."

What This Isn't About

I've told this story in a few keynotes now, and I always have to clarify: this isn't about AI being a better parent than me. (It's not.) It's not about outsourcing parenting decisions to a chatbot. (Please don't.)

It's about something simpler and, honestly, more profound.

In that moment, I was too emotionally activated to access my own good judgment. The part of my brain that knows better—the part that's read the parenting books and believes in connection over punishment—was offline. Hijacked by stress and pressure.

AI gave me access to an objectivity I couldn't reach on my own.

That's the thing that blew my mind. Not that AI knew something I didn't. But that AI could hold space for my better thinking when I temporarily couldn't.

The Implication

I think about this a lot now. We talk about AI for productivity—faster emails, better reports, more efficient workflows. And that's real. (I use work with it for all of that.)

But the broken window moment taught me something different. AI isn't just about doing work faster. It's about being better at being human.

Better at the moments that matter. The difficult conversation you're dreading. The feedback you need to give but don't know how to frame. The decision you're too close to see clearly.

Kevin Kelly (co-founder of WIRED) said on our podcast that AI is "by far the most profound thing the human race has ever done, equal to maybe language or fire." I used to think that was hyperbole. I'm less sure now.

Your Move

I'm not going to give you a five-step framework here. (You know I love a framework, but this one's simpler.)

The next time you're stuck—really stuck, emotionally stuck, not just intellectually stuck—try inviting AI in. Not to decide for you. But to expand the options you can see.

Ask for ten terrible ways to handle it. Or ask it to interview you about what's really going on. Or just... talk to it. Out loud. Like you would a friend.

(If you’re desperate for something to copy paste, think of one decision you're sitting on—something emotional, not analytical. Open ChatGPT and borrow my standard “Terrible Ten” prompt: "Act as [expert]. Give me 10 terrible ways to [handle my situation… (blabber on for a while)]." See what your judgment does with the bad ideas. I bet you'll spot something you couldn't see before.)

You might be surprised what you can access when you're not trying to access it alone.

The first time AI blew my mind, it wasn't at work. It was standing in my hallway, surrounded by broken glass, realizing I'd just become a better parent than I could have been on my own.

That's when I knew this wasn't just about productivity. This was about being human—with a little help.

Related: Malpractice (Invite AI Reflexively)
Related: Lose Your Thumbs, Find Your Voice
Related: The Advisor You Didn't Know You Had

Join over 29,147 creators & leaders who read Methods of the Masters each week.