Hart spent an hour and a half with Microsoft Copilot trying to fix a series of small things on his website.
Copilot wasn't up to it.
He switched to ChatGPT. Fifteen minutes later, everything was done.
He told our mastermind group, matter-of-factly: "Not all AI is equal."
I've been saying this for a while, but now I had a perfect example to point at.
The Janitor and the Book
A lot of people use Claude's Opus model for everything. Every task, every request, same model every time. Then they run out of credits and complain the tool doesn't work.
But Opus isn't the right model for everything.
It's like asking the janitor to collate and print an unpublished book. Maybe the janitor has the skills. But they're not who you should be asking.
Different tools for different jobs. That's not a workaround — that's the whole operating model.
Hart also mentioned, with the restraint of someone who chose not to be angry at a machine, that Copilot had surprise-billed him $77 for the month. He got his bank to reverse the charge.
"I try not to demonstrate anger," he said. "For one thing it's futile. For another, I'm talking to a machine. And for another, that's just giving them everything they need to psychologically profile me."
I laughed. Still laughing.
The Keynote I Keep Coming Back To
We were at a conference recently where Christopher Lockhead gave a keynote. He wrote Category Pirates. His argument: building a slightly better version of something that already exists is the slowest path to winning anything.
He used Pepsi as the example.
Pepsi's entire strategy for years was "we're better than Coke." Taste tests. Blind comparisons. Every campaign built around the contrast.
Guess what happened? People kept buying Coke.
Because when you constantly compare yourself to the category king, you're borrowing their gravity. You're reminding everyone they exist. Coke gets 76% of the market. Everyone else fights over the remaining 24%.
But if you create an entirely new category — define a problem nobody else has named, own a space where no one's planted a flag — you become the reference point. You get to be the one with 76%.
Smart Passive Income is a category. Pat Flynn didn't launch passive income blog number 327. He named a specific thing, branded it, and became the gravitational center everything else orbits.
Live in the Problem
Then Lockhead said the thing I can't stop thinking about.
The mistake most creators and entrepreneurs make is that they fall in love with their solution. They get so excited about what they've built that they pull back from the problem. They declare themselves ready to scale and shift from listening mode to broadcasting mode.
And when you stop being inside the problem, you stop seeing what you still don't understand.
His advice: live in the problem. Not just at the start. Always.
I was supposed to launch PodGlue 25 days before that mastermind call. I didn't do it.
I told the group honestly: I chickened out.
But then I walked through what happened in those 25 days instead.
A host from the Legal Podcast Network heard about PodGlue and reached out immediately. A woman who'd met me once at Podcast Movement, had been following the show for two years, booked a call the day she got my email. A third conversation opened a potential partnership I hadn't seen coming.
None of that would have happened if I'd gone into launch mode.
Launch mode means you switch from receiving to broadcasting. You stop taking in new information. You stop being inside the problem. And when you stop being inside the problem, you stop seeing the parts of your solution that still need work.
Staying in the problem 25 more days gave me three useful conversations and two potential partnerships.
I'm glad the launch slipped.
The Category Nobody's Owning Yet
Shelly teaches piano. She runs a podcast. She said something on the call I keep returning to.
She's noticed that the content ecosystem for piano teachers is crowded — courses, communities, certifications, business coaching. If you teach piano, there's no shortage of people trying to serve you.
But piano parents? Almost nothing.
She's been sitting with parents in recitals for years. She hears the same anxieties over and over — what to do when a kid wants to quit, how to think about practice, what progress actually looks like. She knows this problem from the inside.
But she's been watching it the way I watched the grocery store problem in 2012. Clearly. With full awareness that the gap exists. Without planting her flag in it.
I asked her: have you had this conversation with ChatGPT? Not to get the answer — just to say out loud everything you already know about this problem, and let the tool reflect it back.
Because here's what I've found: the AI doesn't have our limitations. We look at a market gap and immediately calculate reasons it might not work. The model just sees the opportunity and starts building the case for it.
Sometimes you need something that doesn't know you're supposed to be afraid.
Wendy's Seven Email Addresses
One more thing from the call that deserves its own moment.
Wendy spoke at a conference last week. Small room, maybe twelve people. At the end of her talk, she put up a QR code for a free download — something directly related to what she'd just presented.
Seven of the twelve people gave her their email.
Seven out of twelve.
Here's why it works: people at conferences are taking notes. They're worried about forgetting something important. When you say "you don't have to write all of this down — I've got the highlights in a free download" — you're solving a problem they're actively experiencing, right there in the room.
The QR code captures the email at the highest-attention moment of the entire event.
The lead magnet isn't separate from the talk. It is the talk, made portable.
Hart's speaking at two sessions next weekend. He's building a PDF before he goes. The room will be scientists, policymakers, and potential strategic partners.
Even one real connection from a room like that is worth the hour it takes to make a one-page download.
The whole call was like this.
Small group. Real work. Real numbers. Real problems.
That's worth more than most webinars I've sat through.
Stay in the problem. Stay in the room.