The AI Hat Podcast

10/10/2026

When Taylor Swift lost the masters to her first six albums, she recorded them all again. She’d written the songs, so nobody could stop her singing them.

Custom GPTs and Gemini Gems are about to teach a lot of small businesses the same lesson. OpenAI retires custom GPTs on December 11, and Google starts turning Gems into skills in November. Whatever instructions you wrote come with you, as long as you have them written down somewhere you control.

If your team leans on a few of these, the first issue of my new LinkedIn newsletter covers which one to move first, how to test it, and where to keep the file so the next move is an upload instead of a rebuild.

💫

Hello my friend, When Taylor Swift lost the rights to the original recordings of her first six albums, she recorded them again. The songs were always hers.

09/25/2026

Speaker friends, this one's for you. 🎤

Your talk got applause. Then: radio silence.

Think about how much you share every month - keynotes, webinars, summits, coaching calls. Almost all of it gets recorded, and almost none of it turns into content. I'm as guilty as anyone.

That's why I love what Archie does. Give it the transcript from your latest talk, and it pulls out the best angles, drafts posts in your voice and schedules them on your most important channels. Your ideas keep working long after you leave the stage.

Quick tip: if the organizer has your recording, ask for it. It's content you already made.

The link to try Archie free is in the first comment.

09/22/2026

Most AI writing tools have never seen a single thing you've published. They write about your topic instead of from your expertise, and they have no idea which of your posts actually performed.

Archie starts from content you already made - a webinar, an article, a podcast - and ends in your publishing calendar instead of your clipboard. Your Playbook holds the audience, goals and voice, so draft one already sounds like you.

09/13/2026

She spent the whole episode on what breaks an operation, then closed on this.

Every decision that got you here was made by someone doing their best with what they had. That is not a failure worth carrying.

Ty Givens, CX Collective, on episode 124 of The AI Hat Podcast.

09/06/2026

Mike didn't design his own CRM.

He told Claude what he needed to track, what he actually does with that information on a Tuesday, and what decision he wanted to be able to make from it. Claude proposed a structure, he pushed back on about a third of it, and then it built the databases, the properties, the relations between them and the views.

Same with the content pipeline. Same with the tracker holding his trade bylines.

The trick is one clause most people leave out of the request: ask what it would leave out, and why. Skip that and you get every property anyone could conceivably want, in a database nobody maintains past week three.

Your AI starts every chat at zero. Here is how I use Notion MCP to give my Digital Crew a permanent memory - the setup, and the step most people skip.

08/31/2026

The first question in almost every insurance workshop I run is some version of: am I even allowed to do this?

Fair question, and the honest answer shaped the whole book. Fifty-six prompts, and not one of them ever needs to see a client.

Plan documents are public. Your notes about your own week are yours. The client's file stays out of the AI tool entirely, which is what makes the conversation with your compliance reviewer a short one.

Free digital edition: theaihat.com/aep-prompts

08/28/2026

You're good at this. That's not a compliment, exactly.

You've built the workarounds. You know which portal is slow in the mornings. You keep a text file of plan details because pulling them up takes four clicks.

That competence is real, and it's why nothing changes. Every clever workaround makes a broken design survivable, and a broken design that's survivable never gets fixed. It just gets absorbed. By you.

theaihat.com/podcast

08/21/2026

For about seventy years, if you wanted to call somebody, a person had to physically connect you.

Nobody ever fixed that by buying a nicer switchboard. Bigger boards got built. Faster boards. More holes. Every improvement made the operator better at a job that shouldn't have existed.

Your management system can't talk to the carrier portal. The portal can't talk to the quoting platform. So a licensed professional connects them by hand, one call at a time, all day.

You're the cable.

https://theaihat.pulse.ly/mwzvreywbj

08/14/2026

Two questions, ninety seconds apart. Two completely different answers.

Ask a room of insurance agents which task feels most like data entry instead of selling and almost everyone says updating the CRM. Ask what they stopped to look up in the middle of a client call last week and nobody says the CRM.

The switch that annoys you and the switch that costs you are not the same switch.

Full episode at https://theaihat.pulse.ly/ytskfq8mjw

07/29/2026

Imagine your AI agent deletes the wrong records on a Friday afternoon.

No hacker. No break-in. Just your own automation doing something nobody thought to check it couldn't do.

Now imagine the call where your carrier explains why that isn't covered.

Here's why that call is coming for some leaders. Most cyber insurance works like your homeowner's policy: it knows exactly what to do about a burglar. But when your own AI does the damage, nobody broke in. A policy written for break-ins may have nothing to say about it.

The market has noticed. In January, ISO published standard policy language that lets insurers exclude generative AI claims, and carriers adopted it within weeks. In Delinea's latest survey, 42% of companies said their cyber policies already exclude AI misuse or liability.

But buried in the bad news, the underwriters did us a favor. They spelled out exactly what earns coverage: a written list of what your AI can touch, proof you've tested it, and a name attached to who's watching it.

That's AI governance. The homework most teams skipped on the way to deployment.

You can't ban your way out of Shadow AI. Now you can't insure your way around it either. The only path left is to govern it.

So let me ask: when did you last look at what your team's AI can actually touch?