I was sitting in a booth at a Braum’s in Oklahoma with a friend. We had both been eating hamburgers. We were talking about normal mundane stuff before he asked how I was doing and what I was doing now, and I told him I was in between jobs, working on network engineering, and doing a lot with AI.

He asked me to tell him about that and I started going through different things I’ve built. At some point he said he wished we had four more hours for lunch so we could keep talking about it, and he told me what I knew was impressive and there had to be some way to do something with it. I didn’t believe him. I told him I didn’t know how to do that, I don’t know how to market myself or go out and talk to people about me.

This is all self-taught. I have no credential for AI, no degree, and no years of experience doing this for a company. I have two months of obsessing over AI. That’s what I have. I keep thinking anybody could do this, so what makes me different? Maybe it’s because I’m doing it and they’re not. I don’t know.

I’ve spent those two months building projects and listening to podcasts, trying to educate myself and trying things. AI is really probably the only hobby I’ve had where I’ve asked it itself how to teach me how to use it. I didn’t know how to start. I didn’t know how to do any of that stuff. I just started asking AI.

One of those projects was the fantasy football platform I built for my keeper league. In the beginning I told Claude generally what I wanted and gave it a lot of creative license. When I went into the league there were no commissioner tools, no place for me to see who I had invited, who had actually signed up, or what their email addresses were. Claude didn’t know what was in my head. I had it build the list I needed, and then I realized the league had no way to send anything out, so I built an email server on my cluster instead of making and managing another Google account.

I use both Claude and Codex. They can do the programming and a lot of the design, but they still get things wrong. I have to redirect them, steer them, tell them this part was okay, keep this, go do that. I’m probably the project manager. I’m also the customer. Sometimes I’m just some guy repeating, “That’s not what I meant.”

The part that feels a little like having a superpower is that when I realize I need something, I can try to build it. Before, I would have had to find somebody else’s version, pay for it, or just go without. Now my first thought is, well, let’s see if I can make it.

Everything so far I’ve built for myself. I think I have to build something at a production level and have a bunch of people use it and break it before I can figure out, oh, this is what I need to build into it so it doesn’t break. I’ve never been at that level. I don’t know what production-level work requires because I don’t know what I don’t know.

In my mind right now I have the confidence that I could probably build pretty much anything another solo builder could build, unless it needs an enterprise team. Asking somebody to pay me for it is where I don’t have that confidence. I have two months of doing this for myself and no experience doing it for somebody else, and I don’t know how to market myself.

My friend didn’t look at it the same way I did. He wanted four more hours and thought somebody could use what I knew. I saw myself a little differently after lunch too. I’m still not there, but I’m willing to look into it and willing to try.

If we have that lunch again in six months, I want to be able to tell him I built something from it, got better, and figured out how to make it part of my living. I’ve had a lot of hobbies. This is the only one I really wanted to make a career.