Every era of the internet had one big verb.

Web 1 was look it up. You typed something in, a page came back, you read it. The internet was a giant library.

Web 2 was interact. We got feeds, comments, likes, profiles, marketplaces. But all of it happened on fixed platforms someone else built and owned. You could post anything you wanted, as long as it fit inside their box.

Web 3 was move money. For the first time, value could travel online the way information does. Wallets, stablecoins, settlement in seconds instead of days, no bank asking you why.

I think we just walked into the next one. Web 4 is build. People creating the systems that solve their own problems, and those systems talking to each other.

The audience of one

For my whole career, software was something you made for a market. You needed thousands of users to justify building anything. If your problem was too specific, too weird, too personal, nobody was going to write the app. You adapted to the software, not the other way around.

That rule is gone.

I wanted a better version of f.lux on my Mac. So I built one. It's a little menu bar app with sliders that pull blue and green out of the screen, and presets I named Warm, Sunset and Candle. It does exactly what I want and nothing else. No account. No upsell. No ads.

I wanted to see my money the way I think about it. So I built personal finance apps for my desktop and my iPhone. They exist only for me. They will never be in an app store. They don't need to be.

This is what I call the personal software era. Software for an audience of one. A year ago that sounded like a hobby. Now it's just how I solve problems. When something annoys me, my first thought isn't "what app should I download?" It's "what should I build?"

Systems, not apps

Here's where it gets interesting. One personal app is nice. A bunch of personal systems that work together is a different thing.

I run a team of about 30 AI bots. Each one owns a job. One handles finance. One handles taxes. One drafts content. One hunts for work. One manages my inbox. They don't sit in separate tabs waiting for me to remember them. They coordinate, hand work to each other, and report up to a chief of staff bot that tells me what actually needs my attention.

None of that came from a platform. Nobody sold me "Ryan's operating system." I put it together piece by piece, around my life, the way I'd wire a house.

That's the second half of Web 4. It's not just that people build. It's that the things they build interact. My finance system knows what my tax system needs. My content system knows what I'm working on. The output of one is the input of another.

Web 2 gave you one big platform and asked you to fit inside it. Web 4 gives you a pile of small systems that fit around you.

The house is a computer now

The clearest place to see this is a home.

I've been doing open source home automation with ESP32s. If you haven't played with one, an ESP32 is a tiny, cheap chip with Wi-Fi built in. Put a sensor on it and it can tell you the temperature, the humidity, whether a door is open, whether someone walked in the room. Put a relay on it and it can turn things on and off.

Wire a few dozen of those into a house, add apps that control everything, and the house stops being a collection of dumb objects. It becomes a system. Lights respond to the time of day. The AC knows the window is open. The lock knows you're home. And because it's open source, there's no monthly subscription, no locked-in vendor, no company deciding next year that your devices are "end of life."

This is a big part of what I'm doing with Byldr now. I go into people's homes and do a full makeover: plan it, install it, build the apps, and show the before and after. Every house is different, so every system is different. That's the point. It's not a product off a shelf. It's a system built for the people who live there.

Where Web 3 fits

Web 3 didn't go away. It became plumbing.

When you have systems acting on your behalf, at some point they need to pay for things. Pay a bill. Pay for an API call. Settle with another system. Stablecoins are the rails that make that possible. An agent can't walk into a bank branch, but it can hold a wallet and send dollars in seconds, anywhere in the world.

So the stack looks like this. Web 1 gave us the information. Web 2 taught us how to interact. Web 3 gave us money that moves at internet speed. Web 4 is people using all of it to build their own systems, and letting those systems trade, talk and work together.

What changes

A few things I believe are coming fast:

The default answer to a problem becomes "build it." Not for engineers. For everyone. The cost of making a small tool for yourself is falling toward zero.

Platforms matter less. When you can build exactly what you need, the giant app that sort of does it gets less interesting. The value moves to the pieces that connect: open protocols, open hardware, APIs, wallets.

Your systems become an extension of you. The way I run my day, my money, my house and my work is now encoded in software I control. That's a strange and powerful thing. It means improving my life and improving my systems are the same activity.

The interesting stuff happens between systems. One bot is a tool. Thirty bots that coordinate are a team. One smart switch is a gadget. A house where everything talks is something else entirely.

Start small

If this sounds abstract, start with one annoying thing in your life. The report you rebuild every week. The light you always forget to turn off. The spreadsheet you hate. Build the smallest system that fixes it, just for you.

Then build the second one. Then make them talk.

That's Web 4. Not a new platform. Not a new token. Just people building the systems that solve their own problems, and connecting them.

I'm all in on it.