What happens to a gift economy when the gift gets cheap?
This is part of a series of posts I’m writing as I work my way through older books about programming. I’m not done with this one yet, so treat this as a hypothesis, not a finished take.
I’m most of the way through Eric S. Raymond’s The Cathedral and the Bazaar, and one of his ideas has been bugging me for a week.
Raymond splits cultures into two kinds. Exchange and command cultures run on scarcity: you trade for what’s scarce, or someone with authority tells you to go make it. Gift cultures run on abundance: once the basics are covered, status stops being about what you can pile up and starts being about what you can give away. Open source, he says, is a gift culture. Hackers earn reputation by giving code away for free.
I buy it. It matches what I saw in six years around the bootcamp world and a decade in the Rails community. The people with the most status were almost always the ones who’d given the most away.
But the whole framework rests on one assumption: a gift only earns you status if it cost you something to give. Nobody gets reputation for handing over a penny. You get it for handing over something that was hard to produce. And the hard part was never the code itself. Copying code has been free for decades. The hard part was the labor: the skill and the hours it took to write the thing in the first place.
That’s the scarcity that’s collapsing right now. I can generate a competent library in an afternoon that would’ve cost me a week a few years ago. So can you. So can a college freshman. And when the labor stops being scarce, giving the result away stops signaling much of anything.
So what happens to a gift culture when the gift gets cheap?
My first answer is that it doesn’t die. It moves.
When the artifact is free, the scarce thing becomes the stuff wrapped around the artifact. Taste: knowing which of the ten generated options is the right one. Trust: vouching that a thing is correct and will still be maintained next year. Judgment: knowing what’s even worth building. Context: actually understanding the problem instead of just the solution. “I wrote this” stops being the flex. “I’ll keep this alive,” or “I can tell you which of these is any good,” takes its place. The gift moves up a level. Same culture, more abstract currency.
That answer lines up with something I already believe about AI in general: when the work gets cheap, the value moves up to judgment. Fine. Not surprising.
The second answer is the one I can’t stop chewing on, because Raymond had no reason to see it coming.
His hackers could afford to give code away because they were doing fine. They had jobs, time, stability. The abundance that makes his whole model work isn’t an abundance of code. It’s an abundance enjoyed by the people writing it. He never had to pull those two ideas apart, because in 1999 they came as a set. If you could write good software, you were employed and comfortable, full stop.
AI pulls them apart. You can flood the world with code while the people who used to write it for a living watch the floor drop out from under them. And if that happens, you don’t get more gift-giving. You get less. A gift culture full of people who can’t make rent slides straight back toward exchange and command. You give your work away when you feel secure. You sell it, or you do what you’re told, when you don’t.
So the question I keep landing on isn’t “is code abundant?” Obviously it is. It’s “are the makers abundant?” That’s the number I’d actually watch. And right now I honestly can’t tell you which way it breaks.
Like I said, this is just a hypothesis, and I’m not even finished with the book. But I’m on the lookout now. More when I am.