My best qualification for the AI moment is that I’m a millennial.

I think the thing that qualifies me most for this whole AI moment is that I’m a millennial. I know you’re already annoyed with me for saying that, but I really do believe it, so let me walk you through why.

I was born in 1986. The first computer we had in the house was an Apple IIe, one of those machines that ran off the giant floppy disks the size of a dinner plate. The thing I loved to do on it was make the mostly wonderfully pixelated greeting cards, and I genuinely thought that was the most magical thing in the world. And then as I got older, the technology just kept coming. A lot of it was these silly, single-use, kind-of-wonderful little gadgets. I had this Nickelodeon toy where you’d go into a dark room, hang a little white panel on the wall, hit a zapper, and it would snap your shadow right onto it. I thought that was the coolest thing I’d ever seen. My favorite doll was Stretch Armstrong, and to this day I could not tell you what that man was made of. Everything I had was some goofy, wacky 90s version of an electronic, and I loved all of it. But the whole time, there was this feeling underneath: I wanted these toys to be so much more powerful than they actually were.

There was a watch I bought at the mall when I was about thirteen, old enough to make my own purchase decisions by then. It was this gadgety little thing, and it did one silly, useless job, dumb as all hell. And God, I wanted it to be powerful so badly. I wanted it to actually do something for me. And here’s the part that gets me now: that dumb little watch I cried over is basically the Apple Watch. The thing I was desperate for at thirteen exists now. I just had to wait for the world to catch up to what I already wanted it to be.

The one I think about the most is this electronic diary I had when I was seven or eight. It was a pink little clamshell with a tiny keyboard, very serious business. When you turned it on, it made you set a password. And then I never, ever figured out how to get back into it. Not one single time. It was junk from the moment I bought it, honestly, but I wanted it so badly to be a real electronic journal that I loved it anyway. And that’s the thing about all of these gadgets. Every single one of them was ugly, half-working, and not quite there yet. And somehow, I could always see exactly what it was supposed to become. I was never confused about the potential. The tools just hadn’t caught up to it.

That is basically the story of my entire life with technology. Somebody hands me the ugly, clunky, barely-working version of the future and says here, figure this out, and I get to work. When my parents brought home our very first DVD player, they handed the box to me, eleven or twelve years old, because I was the techie of the house. They said, “It’s called a DVD player, go hook it up.” So I spent about four hours flipping the TV around, plugging things in and pulling them back out, clicking the remote. Eventually the whole family sat down together and watched Independence Day, which I picked because I got to buy our first DVD.

I was one of the original Facebook users at eighteen, back when your school had to have a contract with them, and you needed a .edu email address just to get in the door. It was mostly used to meet people on your dorm floor and “poke” them (???). When ChatGPT first showed up at the end of 2022, I started using it almost right away. Honestly, it was awful, but I could see the magic in it. I could see exactly what it would be if it just worked a little better. The bones were all there.

So here I am. Six months ago, I didn’t even know what a terminal was, and now I’m in one every day, working with Claude Code and building things. It’s raw, ugly, looks like nothing, but the output is everything I’ve ever wanted. Which, if you’ve been following along, is the exact kind of thing I’ve been training for since I was a kid, hanging a little white panel on the wall to catch my own shadow.

Why millennials are the real digital natives

This is why I keep saying that millennials are the real digital natives. Gen Z was born into a world where technology already worked and looked beautiful, where someone had figured it all out before they ever picked it up. That makes them fluent, but it doesn’t make them comfortable with the raw, unfinished version of things. Gen X came to it a little later and a little more carefully. We’re the ones who grew up bending clunky, half-broken gadgets to our will and just trusting they’d eventually turn into something better. And we were right. It took a while, but we were right.

Here’s where I actually want to turn this toward you, though, because I don’t think it’s really about me at all. I’m not a coder. I don’t read code, I don’t write code, and I am not a computer scientist. What I am is a digital native. I know how things are supposed to work and how real people actually use them, and it turns out that, in this particular moment, that is the thing that matters most.

The chatbot you use is the widget watch

So if you’ve been sitting there using Claude or ChatGPT, feeling like you spend half your time trying to get the thing to do what you want and it’s still just a little bit off, and some part of you keeps quietly thinking this could be so much better if I could only figure out how, I want to tell you something. You’re not bad at this. That itch you feel, that sense that there is more here than what you’re getting, that is not a you problem. That is you doing exactly what I’ve been describing this whole time. That is you looking at the widget watch and knowing, somewhere in your gut, that it’s supposed to be the Apple Watch.

Because here’s what almost nobody tells you. The chatbot, the little window you type into to write emails and IG posts, that is the widget watch. It is the dumbest, most single-use version of what this technology actually is. It does one thing at a time; it forgets everything the second you close it; none of your projects talk to each other, so you end up re-explaining yourself over and over. And it starts to feel like the tool is fighting you the whole way. A lot of people are quietly deciding, based on these experiences, that this is just what AI is and it isn’t that great. But that’s a little like deciding cars aren’t for you because the first one you ever sat in wouldn’t start. What you’re frustrated by isn’t the ceiling of this technology, it’s the very bottom floor of it.

What “AI agents” actually are

So let me actually tell you what the top floor looks like, because I think “AI agents” has become one of those phrases everybody nods along to right now without totally knowing what it means. Here is the plain version. Instead of one chatbot sitting in one window, you have a whole team of them, and each one is set up to do a specific job for you. I have one that runs my day like a chief of staff. I have one that writes in my voice, one that keeps an eye on my money, one that does research and comes back with real sources, and one whose entire job is to check the others’ work and catch the mistakes before they ever reach me. They are not separate little chats I have to babysit. They know about each other, they hand work back and forth, and they all run on the same shared memory of me.

And that memory is the whole game. My system knows my businesses, my clients, my writing, my standards, the way I like things done. So I am never starting from a blank page, and I am never re-explaining myself. I point it at something in plain English, the way I’d hand a task to a real assistant, and it goes off and does the actual work. Not a write-up about the work. The work itself.

The part that still feels like science fiction to me is that it gets better on its own. The agents check each other’s work and slowly learn what “good” looks like for me. And here is the part I love most: I built this ritual right into the system, so every Friday, without me lifting a finger, it runs what I call a Kaizen review, which is just a Japanese word for small, steady improvement. On its own, it looks back at the whole week, finds where it got something wrong or made me repeat myself, and fixes those things before the next week even starts. So instead of a tool that is exactly as good on day one hundred as it was on day one, or one that slowly drives you crazier, you have something that is measurably a little sharper every Monday than it was the Friday before. That is the opposite of the chatbot, where you are basically starting over every time you open a new window. And I want to be clear, I do not write a single line of code to run any of this. I just talk to it. That is the part almost nobody believes until they watch it happen.

What it looks like on the other side

I’ll give you two recent examples, not to show off, but because they illustrate the difference between the two versions better than I can explain.

A friend’s mom had an apartment she was renting out, and I couldn’t tell whether the price she was asking was market rate. And the slightly embarrassing truth is that I didn’t even know how you find an apartment anymore. Do I use Craigslist? Zillow? I genuinely had no idea where to even start. So instead of Googling around for an hour and giving up, I had my agents put together a quick little tool and pull in some data from a rental site. Within about ten minutes, I was looking at fifty or so real listings for one- and two-bedroom rentals in the exact neighborhoods and price range I cared about. And not only could I answer the original question, I actually found a few places that really excited me. Ten minutes, start to finish, on a problem I had walked into knowing absolutely nothing about.

The other tool I just finished building is for a guy with a Hot Wheels collection, which, up until recently, was a room full of bins with no real way to know what any of it was worth. Now the whole thing runs off his phone. He can log a car in a couple of seconds, and it connects straight to his Shopify store. Then it tells him what each one is trending at on eBay, so he knows which ones to sell and which ones to hold on to. And the thing that gets me is that I could build that same system for a collector of anything, of literally anything, by tomorrow.

That’s the part that still stops me in my tracks. The internet used to feel like a big pile of separate websites that I clicked around between. Now it feels like a library, and I finally know how to walk the stacks and actually pull what I need off the shelf. It’s like I spent my whole life seeing a thousand separate trees, and one day I started seeing the whole forest and all the paths running through it.

And honestly, the reason I’ve gotten a little obsessed, maybe even a little desperate, to teach this to other people is that I don’t want you stuck typing into the widget watch, wrestling with it, deciding this is all there is. Because it really, really isn’t.

Here’s what I’m doing about it

I run this as a small cohort, ten women, and over four weeks I take you from the chatbot version of all this to the actual system version. And I don’t just mean I’ll teach you about it. I mean I am going to give you the entire system I built, the whole operating system, the agents, the memory, the Kaizen ritual, all of it. I have already done the hardest part. I built the brain. It just needs to get filled up with you now, with your businesses, your voice, your clients, and the way you like things done.

In the first week we do real things together, the apartment search kind of thing, so you can feel in your own hands what this is capable of. And then we start pouring you into the system itself, so that by the end you are not watching me run mine, you are running your own.

It’s really for you if you’ve already gotten comfortable past the basic chatbot stuff, if you’ve spent real time in Projects or in Claude Cowork, and you’re brand new to the terminal and Claude Code side of things. You’ll want two screens to work across and a Claude Max plan, ideally the 5x or the 20x, since that is what gets you the best results. And one honest caveat: for right now you’ll need to be on a Mac, because that is the only version I have built so far. The cohort is $2,000.

And if you just read all of that and something in you went wait, is that me, then it probably is. If you want in on the next cohort, get in touch and tell me what you would build first, if you could build anything at all. That is honestly the only prerequisite I care about.