In 2026 I built an AI install business. Real clients, real money, a model that worked. Then I walked away from scaling it.
Every sale was my hours. Discovery call, install call, handover. The product was me, with better tools. Fifteen years of building and exiting online businesses, and I'd quietly built myself another job.
That's the trap. And almost everyone selling AI right now is selling you a version of it.
The pitch is always more. Write faster. Post more. Automate the busywork so you have room for extra work. Faster output, more content, you still in the middle of everything. Just busier.
The point of AI isn't to help you do the work faster. The point is that you stop doing the work.
Services are being dismantled. Agencies, consultants, freelancers, the whole bill-by-the-hour middle. Not because the work isn't valuable. Because that work is exactly what AI eats first. Salaried roles are on the same conveyor. A job is a bundle of tasks, and the spreadsheet only needs enough of them.
What survives is ownership. Human-led brands with real products, real audiences, real positions. People who own the system instead of being the system.
Ownership has two floors. The first: own the system instead of being the system. Workers take the jobs, you keep the judgment. The second is literal, holding things that need no system at all, and it is the floor where I have least standing to teach. Held assets mostly don't pay you until you sell, which is wealth rather than salary, and I have not run that stage far enough to sell you a playbook for it. So there is a pack of tools and an honest gap where a flagship would be.
You can wait for that to happen to you, or you can run the dismantling yourself. Make yourself redundant before AI does. That's not a threat. It's the whole opportunity.
Redundant publishes AI workers. A worker is not a tool. A tool wants your hands. A worker takes a job.
You configure it once. It runs on a schedule or when you ask. It watches your market, scans deal flow, reads your portfolio's news, pressure-tests a decision before money moves. You read the output and decide.
The deciding stays yours. The doing doesn't.
A pack is a crew of workers for one discipline. Five run a business, and one is for the money it makes. A playbook is a complete system from a model I actually ran. There are three: build a service business, productize the one you have, or buy one instead.
Poker taught me more about running businesses than most of what I've read. The lesson is variance. You can play a hand correctly and lose. You can play it badly and win.
Operators who can't tell the difference abandon good strategies after bad luck and double down on bad ones after a lucky win. Decision quality over outcomes. Game selection over grinding. The whole catalog is built on this, and one pack teaches it outright.
No calls. No onboarding. No retainers, no agency, no upsell mid-checkout. You buy a pack or a playbook, install it yourself in ten minutes, and own it.
If that sounds cold, you're not the customer. If it sounds like relief, you are.
The name is also the project. I run newsletters and hold positions, and I'm making myself redundant from my own businesses with the same workers I sell. When something works, it becomes a product. When it doesn't, you'll read about that too.
Own the work. Don't do it.
Weekly. Making myself redundant from my own businesses, in public. New packs announced here first.