Heaven, Hell, or Just a Nicer Rome?

What we're really building when the machines finally clock in

Picture a career counselor cornering a hunter-gatherer forty thousand years ago. "So," she says, clipboard in hand, "what's your passion?" The hunter-gatherer looks at her the way you'd look at your smartphone if it asked about your feelings. Passion wasn't on the menu. The menu was: find food before the food finds you, or before someone faster finds it first. For most of human history, "what do you want to be when you grow up" would have gotten you a short, literal answer. Alive. Ideally.

Once we settled into cities, towns, and villages, this didn't go away, it just got a filing system. You were a farmer, a carpenter, a stone cutter, a priest, or, if the dice rolled badly, someone else's property doing someone else's menial work. And that role wasn't a job in the modern sense, the kind you quit for a better offer. It was you. It got stitched into your family name, your neighborhood, your marriage prospects, and in a great many societies, into a caste system that made sure your grandchildren inherited the same slot whether they wanted it or not. Occupation wasn't what you did between nine and five. It was the answer to "who are you," full stop, and it stayed the answer for most of your waking life, for most of your life's length, across generations.

It's worth pausing on how brutal that filing system actually was, because we like to romanticize the pre-industrial "artisan economy" as a golden age of proud craftsmen. In truth, a huge share of the necessary work (mining, hauling, tanning, the unglamorous end of agriculture) was done by people with no say in the matter at all: slaves, bonded labor, the lowest rungs of whatever caste ladder a society had built. India built one of the more elaborate versions of this, a jati system precise enough to tell a family not just what they did but whom they could marry and which well they could drink from. The stone cutter who could at least pass his trade to his son was, relatively speaking, one of the lucky ones. Most of history's necessary, backbreaking work was done by people history didn't bother naming. That's the foundation this whole conversation about "purpose" sits on, and it's worth remembering before anyone gets too dreamy about the old days when everyone supposedly knew their place.

Then there was the other one percent: the people who never had to answer the food question at all. Aristocracies, priesthoods, the occasional lucky merchant dynasty. Freed from survival labor, they had something the stone cutter never got: time. Blank, unaccounted for, nobody-asking-you-to-justify-it time. Some of them turned it into the Parthenon, into Euclid's geometry, into the Medici funding Michelangelo, into the whole intellectual apparatus we now call civilization. Philosophy, art, and science are, in no small part, hobbies of people who didn't have to farm. Not a flattering origin story. But an honest one.

And some of that same leisure class turned their freedom into an extremely long, extremely well-funded party instead. Rome kept its idle patricians happy, and its restless plebs quiet, with the same trick: bread and circuses, until the circus more or less ate the empire. Every civilization that produced a class exempt from survival work also produced, right next door, a case study in what happens when that class runs out of ideas and settles for entertaining itself into oblivion. Leisure isn't automatically virtuous. It's a coin flip, and the house doesn't always win.

Now here's the strange part, the one that's easy to read past in your outline: that ratio barely moved for centuries. Steam engines showed up, then assembly lines, then computers, and menial jobs kept getting automated away in wave after wave. Every time, someone confidently announced that mass leisure was around the corner. John Maynard Keynes predicted, in 1930, that his grandchildren would be working fifteen-hour weeks by now, filling the rest of the week with what he called "the art of life": wine, philosophy, long walks, that sort of thing. Instead we got the smartphone, the reply-all chain, and a culture that treats being permanently reachable as a virtue. Technology kept freeing up time, and humanity kept finding new work to pour into it, the way water finds a container exactly its own size. The share of people working for survival, against the sliver pursuing art or science for its own sake, barely moved. We ended up with sharper tools. Not more Athenians.

Which is what makes this AI moment different, or at least worth taking seriously as different. Unlike the steam engine, artificial intelligence isn't proposing to automate one category of task at a time. It's proposing to push the cost of food, health, education, and a great deal of physical and cognitive labor toward something close to zero, all at once. If that happens even partially, the old math breaks for the first time in recorded history. Not a wave of automation moving people sideways into new jobs, but a real shrinking of the total survival work available to justify anyone's calendar. The stone cutter's job doesn't migrate to a new task. It simply stops being necessary. Multiply that across enough professions and the aristocratic condition, time without an economic excuse, stops being a rare inheritance and starts looking like everybody's problem.

Notice I called it a problem, not a gift. Two things tend to get quietly skipped when people wax lyrical about a coming age of AI leisure.

The first is who ends up owning the machines. Every leisure class in history got its free time by sitting on top of somebody else's labor or somebody else's land. If AI-driven abundance arrives and the compute, the models, and the resulting wealth pool mainly with a handful of platforms and their shareholders, we haven't actually dissolved the aristocracy-versus-everyone-else structure. We've just automated the peasants out of the frame and kept the aristocrats. A world where a small number of people own the intelligence running everything, and everyone else gets a subsistence stipend and an ocean of unstructured time, isn't obviously heaven. It might just be Rome with better broadband. Who captures the abundance matters at least as much as whether the abundance shows up.

The second thing that gets skipped is quieter and more personal: most people, handed the aristocrat's blank calendar without the aristocrat's centuries of cultural training in what to do with it, do not spontaneously turn into philosophers. We've actually run small trials of this. Lottery winners as a group do not reliably become patrons of the arts; plenty report the windfall made their lives worse, not better. Towns that lost their defining industry overnight, a mine, a mill, didn't pivot en masse into painting and civic theatre; many are still dealing with the wreckage decades later. Involuntary retirees describe a familiar arc: relief, then a slow, disorienting loss of the structure that used to give a day its shape. None of this is destiny. But it suggests that handing someone time is not the same act as handing them purpose, and confusing the two is how a society ends up with abundant leisure alongside record loneliness and doomscrolling as a full-time hobby. We didn't even need an emperor to fund our version of the circus. We built it ourselves, for free, in our own time, which might be the more embarrassing failure of the two.

There's an older, stranger precedent worth sitting with here, and it isn't the Roman one. Around the same few centuries, roughly the middle of the first millennium BCE, agricultural surplus freed up enough people across several unconnected civilizations, Greece, India, China, Persia, that historians sometimes call it the Axial Age: the era that produced the Buddha, Confucius, the Hebrew prophets, and Greek philosophy, all more or less at once, all as byproducts of societies that suddenly had people with time to think and no immediate famine to think about. Nobody fully agrees on why it happened in parallel across unconnected civilizations, only that it did. Surplus, in other words, has produced both Rome's circuses and civilization's deepest religious and philosophical traditions, sometimes from the very same societies. Which one you get seems to depend less on the surplus itself and more on whatever cultural scaffolding happens to be standing when the surplus arrives.

Which is really the only useful question left, and it isn't a technical one. If we're heading toward an Axial Age moment rather than another Roman one, a few things seem worth putting on the table, not as a program, more as a starting list for the argument. One is ownership: whether some of the gains from AI infrastructure get structured as a shared claim, closer to a sovereign fund with dividends than a purely private windfall, so that the freed-up time comes with the freed-up means to use it. Another is education, and specifically whether we keep training children almost exclusively for employability, or whether some real portion of that effort shifts toward cultivating craft, curiosity, and aesthetic judgment as ends in themselves, the way a guru-shishya tradition or an old-fashioned apprenticeship once did, deliberately, rather than as an accident of aristocratic upbringing. A third is cultural: whether we treat the coming surplus of time as a design problem worth arguing about in public, the way we currently argue about growth or inflation, instead of assuming it will sort itself out because the technology is exciting.

None of that is a policy paper, and it isn't meant to be one. It's closer to a starting list for a dinner-table argument that most people haven't started having yet, because the technology still gets discussed as a productivity story rather than the civilizational one it actually is. Every generation handed unearned time, the Athenian aristocrat, the Renaissance patron, the involuntary retiree, the lottery winner, has already run this experiment for us, in miniature, with famously mixed results. AI is about to run it again, at the scale of the entire species, all at once, with no dress rehearsal and no one else to blame if it goes wrong. Rome at least had an emperor to blame for the circuses. We will only have ourselves, and a scroll bar.

The machines, for the first time in history, are plausibly willing to foot the bill for the shift. What they cannot do is answer the question one economist saw coming a century early, and named more precisely than anyone has managed since.

"Thus for the first time since his creation man will be faced with his real, his permanent problem: how to use his freedom from pressing economic cares, how to occupy the leisure which science and compound interest will have won for him, to live wisely and agreeably and well."

— John Maynard Keynes, Economic Possibilities for our Grandchildren, 1930