Nate B. Jones, an analyst who writes about how AI is changing software, published a video in August 2026 called Nobody Laid Out The Five Kinds Of Software You Can Make. So I Did. In it, he describes a category he calls personal software: tools built by one person, for one person, to solve a problem that would never justify a startup. His examples are ordinary. A couple builds a small app to track maintenance on their house. Someone in Amsterdam builds a tool that tells them when the next ferry is arriving so they can time the walk from home.
When I first heard these examples, my instinct was to treat them as charming edge cases. Then I sat with it longer, and I think Jones is pointing at something bigger than he lets on.
Personal software looks like a small idea. It is actually the first place where ordinary people can own the productive layer of their own lives.
The tool is not the point
If you ask what the couple actually built, the answer is boring. It is probably a form, a list, a few reminders, maybe a chart. Any competent model can now generate the code in an afternoon. If they lose the app tomorrow, they can ask a different model to build it again next week.
So the code is not what they own. What they own is everything that took real time to figure out. Which parts of the house need attention every season. Which contractor did the roof and what he charged. Which appliance is under warranty until when. The rules they slowly worked out for what counts as urgent and what can wait until spring.
That is the part no model can regenerate. It is specific to their house, their money, their patience. Jones puts the same idea plainly when he says the freedom comes when you own the project files, the data, and the decisions.
Generating the app is cheap. The specifications, records, and decisions accumulated around it are the part that stays valuable.
The context is the moat
I keep coming back to this because it flips the usual instinct. For years, the advice was to own the software. Now the software is easy. What is hard, and getting harder to replace, is the context you build up while using it.
Think about what happens over a year. The house app collects a maintenance history. The ferry tool learns which routes actually match the owner's calendar and walking pace. A small budgeting tool learns which categories the household actually argues about. None of this exists inside the model. It exists in the files the person keeps.
That accumulation is the differentiation moat over time. Two people can start with the same generated app on day one, and after twelve months of use, their versions are not comparable anymore. One has a rich record of a specific life. The other has an empty shell. The gap did not come from better code. It came from the person patiently feeding real context into something they control.
I wrote about a related idea in Your Real Asset Moved Upstream: the durable asset is not the artifact, it is the specification and intent behind it. Personal software makes that visible at household scale.
The moat is not the tool. It is the specifications, history, and decisions that only accumulate when you use the tool as yourself, on your life.
What ownership actually means here
Ownership is a word that gets thrown around, so it helps to make it concrete. Ask a simple question: if the model provider disappeared tonight, or doubled the price, or changed the terms, what could the person walk away with?
If the answer is a folder of plain files they can read, export, and hand to another model or another tool, they own something real. If the answer is a login to somebody else's cloud where their data lives in a format only that vendor understands, they own nothing. They rented an experience and called it theirs.
This is why I keep pushing back on the framing that personal software is just a smaller SaaS product. A SaaS product asks the user to hand over their context in exchange for convenience. Personal software, done honestly, lets the user keep the context and treat the software as replaceable.
I made a version of this argument for enterprise tools in When Software Becomes Disposable, Trust Becomes Luxury. The same logic applies at the kitchen table. When the tool is cheap and the context is precious, the person who keeps the context in a form they can actually read and move is the one who ends up in control.
Ownership means you can leave with your files intact and rebuild the tool somewhere else. Anything less is a rental with your data as collateral.
Why this matters beyond one household
The reason I care about ordinary examples like a house app or a ferry tool is that they show, at a scale a person can actually feel, what it means to own the productive layer of your own life. Not to sell it. Not to turn it into a company. Just to have it, keep it, and let it grow with you.
For most of the last decade, the productive layer of daily life has quietly migrated into other people's systems. Your notes, your history, your patterns, your preferences. AI now makes it economically reasonable to bring some of that back and keep it under your own roof, in files you can read.
That is a quieter shift than most AI stories, but I think it is the important one. The democratic effect of this technology is not that more people can write code. It is that more people can, if they choose to, own the specific context of their own lives instead of leasing it back from whoever built the app.
The point of personal software is not smaller businesses. It is ordinary people keeping the specifications and history of their own lives in their own hands.
