Labor Is the Cheap Part

A few years ago, putting together a working demo took me weeks. Now it takes a few days, sometimes an afternoon. The market research demo on this site took me a few days to build, and the thing it does, assembling a full study on a business from just a type and an address, is weeks of manual digging if a person does it by hand. As a senior engineer, what that really means is that real needs, actual problems people have, now fit inside my hobby time. So I keep building them. And I keep giving them away for free.

I have a hard time charging for this stuff, even though the pull to is definitely there. I keep seeing these quick cash-grab sites that promise a big result and, honestly, deliver it. But the effort that went into building them is about the same as the effort it took me to describe the problem in the first place. I don't think there's anything wrong with charging for that. I just have a hard time doing it myself. What I'd rather do is put a tool in front of you that might actually help, and get you thinking about how you'd build your own. That's worth more to me than the tool is.

Because here's the thing people keep saying when they see one of these fast tools: how can humanity survive this. They watch an app get written in minutes that looks like the job they've done for years, and they figure their whole position just got erased. I get it, and some of that fear is fair. But I think it's aimed at the wrong part. What got replaced is the labor: the typing, the grunt work, the tedious middle. The thinking, the analysis, deciding what the result even means and what to do with it, that's still yours. An app can spit out the data. It can't tell you the data was the wrong data.

And past all of that, the hard problems are still hard. There are companies out there right now throwing everything they have at them, and they are not running out. AI speeds up the research part, sure. But figuring out how to read the data, how to explain it to people who have to act on it, that is still deeply human. Making the data easier to get doesn't make it speak for itself.

So when I build a tool that makes some data easier to get and then give it away, I'm not closing a door on anybody. I'm moving the start line up. And if you're the one using the tool, I think the move is to ask three questions:

• How can I use this?
• How could someone else use this?
• Because this exists, what can I build now?

That's the part the fear misses. When the labor gets cheap, the thinking doesn't shrink to match. It grows. There's more to point it at, not less. The demo is free, the framework I built it with is on GitHub, and these essays are me giving away the thinking, which is the part I actually care about. Take any of it. If it gets you building something, that's the whole point.