Mar 22, 2026 Mar 22, 2026 2026-03-22

My favorite moment of The Wizard of Oz is the ending.

This is where Dorothy’s long trip ends – in the city of Oz. She’s survived a tornado, a dark forest, flying monkeys, a witch with a very specific grudge. She’s made friends who each carry their own quiet sadness. She’s walked a long road, literally, and at the end of it there is supposed to be someone with answers. Someone with power. Someone who can finally fix things for her.

And there he is: the Wizard.

At first, he delivers. There’s fire and smoke and a voice that fills the room and a face the size of a wall. You don’t question it. You just believe, because everything in the room is designed to make you believe, and because you’ve come so far and you need it to be true.

Then a small dog wanders behind a curtain, and the whole thing comes apart.

Behind the curtain there’s just a simple man. A nervous, ordinary man surrounded by levers and buttons, doing his best to keep the performance from collapsing. He’s not evil per se. He’s not even particularly deceptive, in his heart. He’s just someone who discovered he could make a very convincing impression of power, and then found himself trapped inside that impression, unable to get out.

What’s strange is that this moment, which should feel devastating, somehow feels like a relief. Oh, it’s just a person. Just someone figuring it out as they go. Just like us.

I think about that room behind the curtain a lot lately.

Because AI just handed one to everyone.

The Billion-Dollar One-Person Company

There’s an idea that’s been circulating in certain corners of the internet, spoken with the particular excitement of someone who has just discovered something that feels like a cheat code.

The idea is this: with the right combination of AI tools and agents, a single person could run what used to require a company. Not a small company. A serious one. Code gets written, support gets answered, marketing gets made, operations get handled, and somewhere in the middle of all of it, one human being sits at the center and directs the whole orchestra. A billion dollars in valuation. One person. The final leverage play.

And the reason this idea won’t die easily is that it’s not entirely wrong. Something genuinely shifted in the last couple of years. The output side of the equation changed in ways that still feel a little unreal. Things that used to require a team of twenty, then ten, then five, can now be done by one person with the right setup and enough patience to learn the tools. Apps get built overnight. Businesses get started. People who couldn’t have launched something five years ago are launching things now, and some of them are actually working.

But the pitch, like all the best pitches, leaves something important out.

The levers don’t pull themselves behind the curtain.

The Man In The Backroom

Someone still has to stand back there. Someone has to decide when the AI output is actually good and when it’s confidently, subtly wrong. Someone has to handle the thing that broke in the night, the customer who is genuinely upset, the strategic question that no model can answer because it requires knowing what you actually value and what you’re willing to risk.

Someone has to hold the whole shape of the business in their head, all the time, and know which lever connects to what.

That person is you. And that person, no matter how good your tools are, doesn’t scale the way the output does.

This is the part of the conversation that tends to get skipped over, because it’s less exciting than the cheat code version of the story. But it’s the part that matters most, especially if you’re building something you actually want to sustain.

The cognitive load of running a one-person operation with AI leverage isn’t lighter than running a small team. In some ways it sits heavier, because there’s no one beside you to say wait, are we sure about this? No one who shares your mental model, your values, your go-to strategies. No one who loses sleep over the same things you do.

You become a single point of failure dressed up in the appearance of something larger, and that appearance, over time, starts to cost something.

The Wizard of Oz had an entire city believing in him, and he was quietly unraveling behind a wall. That’s not a moral failing. That’s just what happens when the distance between what you show the world and what you actually are gets too wide, and stays that way for too long.

When Toto Shows Up

The thing about curtains is that they eventually move.

Not always dramatically. Sometimes it’s small. A customer gets a response with your name on it that was generated by something you didn’t review carefully enough. A decision you handed off to an automated process turns out to have been the wrong decision, and now it’s your problem in a very human and public way. Someone asks a question about how you work and the honest answer is more complicated than the impression you’ve been giving.

The Wizard had no contingency for that moment. The curtain moved and his whole identity went with it, because the identity was the curtain. He had nothing underneath it to stand on.

The better version of this story, the one worth building toward, is knowing that Toto is coming – and building something honest enough to survive it. Not confessing every limitation in every piece of marketing. But keeping enough self-awareness that when someone looks behind the curtain, there’s still a real person there who can say: yes, this is how it works, this is what I actually do, this is what I genuinely provide. And mean it.

The Emerald City Was Still Real

Here’s the part of the story I keep coming back to, though. The generous read.

The Wizard was performing, okay, I got this. But Oz wasn’t fake. People had real lives there. The city functioned, the streets were kept, things got done. Dorothy didn’t get what she thought she was going to get from that room, but she got something. The journey meant something. The friends she made along the way were real.

AI-assisted work can be like that. You don’t have to be as large as you look to create genuine value for real people. A small operation with good leverage can serve customers well, build things that matter, make an honest living. The presentation of scale isn’t automatically dishonest. Sometimes it’s just efficient. Sometimes it’s simply how things work now.

The line gets crossed somewhere else. It gets crossed when the smoke and mirrors start covering up actual failures. When you know something isn’t right and you ship it anyway because the tools made it fast and fast felt like good enough. When the curtain stops being a presentation and starts being a way of hiding from accountability.

The curtain is a tool. The trouble starts when it becomes a crutch, and then slowly, without you quite noticing, your assumed identity.

Don’t Go It Alone If You Don’t Have To

The Wizard didn’t need to be alone behind that curtain. That was a choice – or maybe a trap he walked into gradually, the way most traps work. A small honest crew around him would have changed everything. Someone to share the weight of the performance. Someone who knew what was real. Someone to say, when things got strange, maybe we should just tell them who we actually are.

That’s still the better version of this story, even now. It’s not about “don’t use the tools, don’t use AI because it’s not genuine.” The tools are extraordinary and they’re only going to get stranger and more capable and more woven into everything. But maybe don’t go all the way to one person if what you’re building genuinely deserves more than one person holding it up. Find your Dorothy. Find your Scarecrow. Find someone to stand behind the curtain with you, who knows which levers do what, and who you trust enough to see the whole picture.

The leverage is real. The output is real. The value you can create is real.

Just make sure the person creating it is real too.

Because at the end of the day, no matter how good the smoke looks, no matter how big the voice booms, no matter how convincing the face on the wall — it’s still just you back there, doing your best, hoping the dog doesn’t wander too far.

Know that. Own that. Build from that.

That’s the only version of this that lasts.