Not personal software; Personal companies.

With LLMs it's very fast and fairly easy to build a basic version of almost anything. A habit tracker. A budgeting tool. A custom dashboard for your side project. You can go from idea to working prototype in an afternoon.

But that's just it. It's a basic version.

There's a lot of talk right now about how AI will bring in an age of "personal software." Custom one-off applications built just for you, or maybe for you and a few friends. Tools that fit your exact workflow, your exact preferences, your exact brain.

It's a compelling vision. I just don't think it holds up.

The escalation

I can ask an LLM to build me a simple TODO app. It works. It's local. It does what I need. Great.

Then I want to access it on my phone and my desktop. So now I need accounts, an auth provider, and a database. Fine, I can set that up.

Then I want to see trends. How many tasks am I completing per week? Am I getting better or worse? So I add analytics.

Now it looks ugly to me. I want it to feel good to use. So I spend a day on the design.

Then a friend wants to try it. So I need hosting, a domain, maybe some basic onboarding.

Then another friend wants it. And another. Now I'm thinking about uptime, backups, and edge cases I never considered.

Each step is individually small. But the escalation is predictable. What started as "a quick personal tool" quietly becomes a product. And now I'm maintaining it.

The $5/month reality

At some point, it's just easier to pay someone else.

This is the part the personal software narrative glosses over. The cost of existing apps isn't arbitrary. App stores take a cut. Payment providers take a cut. Hosting costs money. Domains cost money. Support costs time. And the opportunity cost of maintaining your own tool instead of building something else is real.

Most consumer software is already priced near the floor. The $5/month habit tracker exists because someone decided that particular problem was worth solving full time. They handle the auth, the syncing, the backups, the edge cases, the design polish. You just use it.

I can build personal software. I do, a lot. But for most problems, the honest calculus is: my time is better spent elsewhere. Use the free tier of five different crossposting tools instead of building your own. Pay for the $8/month budgeting app instead of maintaining a custom spreadsheet that breaks every quarter.

Personal software works for scratching a very specific itch. It doesn't scale to replacing the dozens of tools most people rely on daily.

What we'll actually see

The real shift isn't personal software. It's personal companies.

LLMs don't just let individuals build tools for themselves. They let small teams build real products that compete with incumbents. What used to require a company of 200 people can now be done by four. The barriers to entry have collapsed, but the barriers to quality haven't disappeared. They've just become achievable by fewer people.

This changes the landscape. You don't need to pay a fortune for an enterprise social media manager when a lean two-person startup offers everything you need at a fraction of the cost. The bloated incumbents with massive teams and massive overhead are suddenly competing with focused, fast-moving alternatives.

The quality and quantity of software will go up. The moats that big brands relied on, like team size, sales force, and marketing budget, will matter less. The apps will survive, just in leaner form. And the new competitors will be everywhere.

Personal companies, personal brands

Instead of everyone building their own private tools, we'll see more people building small focused businesses around problems they actually care about.

A small set of problems. A small group of people who want to improve a specific corner of the world. That's the unlock.

I care about health, habits, and wellness. So I'm building ExistPlan, a suite of tools to live intentionally. It's not personal software. It's a personal company. The difference is that I'm not just solving the problem for myself. I'm solving it well enough that other people want to use it too, and I'm willing to maintain it because the domain matters to me.

This is the pattern I expect to see more of. Not millions of people building private apps that only they use. But thousands of small teams building focused products in niches they understand deeply. Competing on taste, speed, and specialization rather than headcount.

The real opportunity

LLMs didn't create an age where everyone builds their own software. They created an age where anyone can build a company around a problem they care about.

That's a bigger deal. Personal software is a novelty. Personal companies are an economic shift.

Build for yourself when it makes sense. Pay for existing tools when they're good enough. And if you find a problem that no one is solving well, and you actually care about it, build a company around it.

That's where the leverage is.