What Kind of Complexity Should We Afford Ourselves?

Ruth Schwartz Cowan, More Work for Mother (Basic Books, 1983). Still in print, still funny, still right.

There’s a book I keep pressing on people: Ruth Schwartz Cowan’s More Work for Mother, from 1983. It’s a history of household appliances, which is not where you’d expect to find good jokes, and it’s genuinely funny. It’s also one of the most useful things I’ve read for thinking about technology.

The setup is this. The men who invented and sold the vacuum, the washing machine, and the dishwasher pitched them as liberation: machines that would free the housewife from drudgery so she could finally get on with higher things. What actually happened was stranger, and Cowan walks through it appliance by appliance.

The vacuum cleaner didn’t end floor-cleaning. It raised the standard of what counted as a clean floor, so we cleaned more often, to a level no broom had ever been asked to reach. The washing machine made it faintly shameful to wear the same shirt two days running. Laundry went from a weekly ritual the whole household pitched in on to a near-daily job that fell to one person, usually the woman, while the men who used to help quietly exited. We washed more clothes than ever. And the dishwasher, which really did take the dishes off our hands, freed up exactly the time and the plate budget that more elaborate cooking demands. I half-joke that there’s no small-plates restaurant without the dishwasher. Fourteen courses means a pile of crockery nobody would tolerate washing by hand.

So the appliances saved labor and created more of it, at a higher standard, mostly without anyone deciding that on purpose. Cowan’s own summary is blunter than mine:

modern technology enabled the American housewife of 1950 to produce singlehandedly what her counterpart of 1850 needed a staff of three or four to produce: a middle-class standard of health and cleanliness for herself, her spouse, and her children.

What I take from Cowan is that the interesting question about a labor-saving technology is almost never how much labor it saved. It’s what new complexity the saving let us take on. The machine lowers the cost of something, and we spend the savings, on doing more, or on doing something we couldn’t have attempted at all before. You could call the front half of this the Jevons paradox, the old rule that making something cheaper makes us use more of it. But Cowan is after what economics leaves out: not how much we consume, but what we come to expect, and who gets handed the new work.

Which is how I’ve come to think about AI. The efficiency story, the same work done faster and cheaper, is real and worth understanding. It’s also the boring half. The more interesting half is the second- and third-order effects, the ones you can’t see from the starting line. The engineers who put GPS satellites in orbit were thinking about positioning and navigation. They were not thinking about Uber, and they certainly weren’t thinking about surge pricing, an entire pricing economy sitting three layers downstream of a system built to help things find themselves on a map. That’s the half of AI worth watching: the business you could run in a completely different shape, the things that become worth doing once they stop being prohibitively expensive. The question quietly shifts from how much time AI saves to a better one: what kind of complexity do we actually want to afford ourselves with it?

There’s a second reason I like the dishwasher as the reference object, and it’s a little mischievous. If you treat AI as a noble god, a vast new intelligence arriving to either save us or end us, you mostly end up in conversations about AGI and its consequences. Those are fun, and useless to anyone trying to get something done on a Tuesday. Treat it as a dishwasher and the questions turn functional immediately. What does this thing actually do well? What can I stop doing by hand? What could I now attempt that I’d long since filed under too much work? You can’t ask a god what it’s for. You can ask that of an appliance.

Cowan’s housewives never got their afternoons back. They got a different and more complicated domestic life, for better and worse, and almost none of it was chosen on purpose. The complexity is coming with AI the way it came with the dishwasher. The part we get to do differently this time is choose which of it we actually want.