Strategy Sandbox. Status: original concept, first publication, 3 June 2026.

The digital twin is having its moment, and the pitch is seductive: build an AI version of yourself, train it on everything you know, and let people talk to it when you’re not there. A version of you that works while you sleep, that your audience can navigate to, that your grandchildren could still talk to long after you’re gone. It’s a lovely idea, it’s gathering a crowd, and every bit of the crowd’s excitement skips the one fact that matters. You already have a digital version of yourself. You didn’t build it, you didn’t agree to it, and it’s been talking to people for a while now, shaping how the world sees you without you ever knowing.

The twin everyone’s selling is opt-in. You build it, it lives on a platform, and a visitor chooses to go there, which means they already know your name before they arrive. The AI’s version of you is the opposite on every count. The engines built it, not you, from whatever they could gather. I’ve made this case in print before: the moment someone asks an AI about you, it becomes your digital representative. It doesn’t sit on a platform you picked, it sits inside ChatGPT and Gemini and the rest. Nobody navigates to it, because it surfaces on its own, unbidden, in the middle of a conversation about something else entirely. And most of the people who meet it have never heard your name. One is a twin in a room you built. The other is already walking around outside, talking to strangers, with no door and no script.

You don’t have a twin, you have a septuplet, and it’s arguing about you

“Twin” implies a single copy, and a single copy is the comfortable version of the story. The wild gives you one per engine, each trained on a slightly different slice of the record, each telling a slightly different version of you, some of them confusing you with someone else entirely, none of them reconciled with the others. That’s not a twin, it’s a room full of near-copies disagreeing about who you are, and the disagreement is the damage, because an engine that meets contradiction hedges, and an engine that hedges doesn’t recommend. You set out imagining you’d build one twin, and the truth is you already have seven, and they don’t tell the same story.

The most quoted line in the whole twin pitch is train it before anyone sees it, and it assumes a clock that already ran out. There is no before. The seven are out there now, seen, recommending you or hedging on you this minute, and almost nobody has looked at what they actually say. The real work starts the moment you accept that: an audit of the twins that already shipped without your sign-off, which is what it means to read the résumé AI is quietly writing for you.

The rung nobody admiring the twin can see is the close on your own name

Here’s the part the opt-in story can’t reach, because in a walled room it has already happened. When someone types your name into an AI, that person isn’t browsing, they’re at the door, half-sold, and the only question left is whether the machine standing there shows them in or lets them hesitate. That’s not identification, that’s the close. In a built twin, everyone who arrives already chose to come, so the close happened before the twin ever spoke. In the wild, the name-search is the close, happening inside the engine, on a warm lead you didn’t send and can’t see, and your seven twins are either closing it or losing it right now.

Read your wild twins as a salesforce and the three rungs name themselves. At the bottom, the closer takes the warm lead at the door, the one searching your name, and either walks them in or doesn’t. In the middle, the recommender puts you on the shortlist while they’re still comparing. At the top, the advocate carries your name into rooms where nobody asked for you. A version of you that merely gets your facts right is a reliable witness. A version that says the nicest true things to a warm lead is a closer. The facts are the floor, the close is what the rung is for.

Layer	What the AI does	What it runs on	The role it plays for you
Understandability (BOFU)	Identifies you	Facts: who, what, whom	Closer
Credibility (MOFU)	Vouches for you	Proof: independent corroboration	Recommender
Deliverability (TOFU)	Deploys you	Content: granular coverage	Advocate

In a sealed room convergence is free, in the wild it’s the whole achievement

The twin-builders have one thing exactly right, and it’s worth saying plainly: a digital version of you can outlive you, can keep working when you’re absent, can carry what you know to people you’ll never meet, your grandchildren included. For me, that’s true, and it’s worth doing, and it’s also already happening to you whether you’ve built anything or not. The difference is control. Inside the room, the twin is singular because the environment is sealed, so the seven never get a chance to disagree, and convergence comes free. In the wild, singular is the prize, because the only way to collapse seven arguing twins back into one is to feed every engine a version of you so clear and so corroborated that they stop telling different stories. One world hands you convergence behind a wall. The other makes you engineer it out in the open, which is the harder job and the one that actually decides whether the machine sells you or hedges on you. I know the open version holds because it’s been measured from outside: tested against more than 500 AI-visibility professionals across nine models, the method put me in a category of one.

So the next time a digital twin makes you go wow, and it should, the idea is genuinely good, ask the quieter question the wow skips. You already have seven, so do they agree, do they close, and is the version out there talking to strangers one you’d recognise as yourself? The built twin is a room you can show people into. The wild twin is already in every room, and it never asked permission.

---

The distinction between the opt-in, sealed digital twin and the AI’s unbidden representation of you in the wild, the observation that this representation is plural (one per engine) and that convergence is the personal-brand achievement, and the framing of the name-search as the close, are articulated here by Jason Barnard, 3 June 2026.