I came across a story last week that I haven’t been able to shake. Workers in India are strapping smartphones to their heads and recording themselves doing ordinary tasks (cooking, cleaning, folding laundry) so that humanoid robots can learn to do those same jobs. The companies pay them well today. The robots take the work tomorrow. And here’s the part that sticks with me: the workers know exactly what’s happening. They’re doing it anyway, because the money is real right now and the replacement is still a “someday.”
If you’ve been paying attention to where AI and robotics are headed, this probably doesn’t surprise you. But it should still bother you. Because this isn’t a story about one company or one country. It’s a preview of a question none of us have fully answered yet: What happens to an economy (and to a society) when the machines can do the digital work AND the physical work?
I don’t think anyone has a clean answer. But I think we owe it to ourselves to actually sit with the question instead of either panicking or shrugging it off, which is what most people seem to be doing.
The Logic Trap Everyone Falls Into
Here’s the simple version of the fear: robots and AI take the jobs, fewer people earn wages, demand for goods and services collapses because nobody can afford them, and the whole system seizes up. Machines making things nobody can buy.
It’s a tight little logical loop, and it’s not crazy. But it also assumes something that isn’t guaranteed: that the only way people participate in an economy is through a paycheck tied to labor. That’s true today. It has been true for basically all of recorded history. But it’s not a law of physics. It’s a structure we built, and structures can be rebuilt.
The real question isn’t “will machines take over the work.” That part is already happening, and it’s not stopping. The real question is who owns the machines, and whether the value they create gets distributed widely or concentrated narrowly. That’s not a technology question. That’s a political and economic design question, and it’s the one almost nobody is debating loudly enough.
We’ve Actually Done This Before
It’s worth remembering that this isn’t humanity’s first rodeo with mass technological displacement.
Mechanized farming wiped out something like 90% of agricultural jobs in developed economies. The industrial revolution destroyed entire categories of skilled artisan work. Computers eliminated typists, bank tellers, and travel agents by the millions, almost within a single generation. Every single time, the same fear showed up: “This will leave nothing for people to do,” and every single time, new categories of work emerged that nobody could have predicted in advance.
That history doesn’t prove this time will turn out fine. The honest counterargument is that this wave is different; broader, faster, and hitting cognitive and physical labor simultaneously instead of one at a time. That’s a real distinction. But it means the burden of proof is on showing this time is different, not just assuming it because the headlines are scary.
There’s also a strange counterexample worth knowing: Japan. It’s one of the most heavily automated manufacturing economies in the world and has been for decades. It doesn’t have a mass unemployment crisis. It has a labor shortage. Automation there didn’t create joblessness; it absorbed a shrinking workforce. That doesn’t mean the same outcome applies everywhere. But it’s proof that “automation equals mass unemployment” isn’t an iron law either.
Where the Real Danger Is Hiding
If I had to point to the actual structural risk, the thing that should keep people up at night more than “robots are coming” is this: ownership concentration.
If a small number of companies and individuals own the robots, the AI models, and the infrastructure that runs them, then capital compounds endlessly while ordinary income from labor shrinks toward zero. That’s not really a new version of capitalism. That starts to look a lot like a high-tech version of feudalism: a small ownership class and everyone else dependent on whatever arrangement that class is willing to offer.
There’s a second piece of this that doesn’t get talked about enough, and it hits closer to home for a lot of the world than people in wealthy countries realize. The traditional path for developing nations climbing toward prosperity has been cheap labor, manufacturing jobs, skill accumulation, a growing middle class, and eventually a more developed economy. That’s the playbook South Korea followed. Taiwan followed it. India and Vietnam are following it right now.
Humanoid robotics threatens to take that ladder away entirely. If a robot in Ohio can manufacture goods cheaper than a worker in Chennai, with zero shipping delay and no labor negotiation, the entire comparative advantage that gave developing economies their on-ramp disappears. That’s not a “some people lose their jobs” problem. That’s an “entire development strategy that lifted hundreds of millions of people out of poverty may no longer work” problem.
So Is This Actually Going to Be Fine, or Not?
I’m not going to pretend I know. But here’s how I think about the range of outcomes, because “utopia” and “collapse” aren’t the only two options on the table.
The most likely near-term path isn’t collapse; it’s a widening split. A relatively small class of people who own AI and robotics capital does extremely well. A larger group ends up in jobs that are harder to automate (care work, skilled trades, anything requiring genuine human trust or presence) and gets by but watches the gap between themselves and the ownership class grow. That’s not the end of the world, but it’s also not a great world if it goes unmanaged for decades.
The more hopeful path requires governments to actually act: automation taxes, some form of guaranteed income, or real ownership stakes in AI infrastructure for ordinary people, not just shareholders. That’s technically solvable. Whether democratic institutions, anywhere, are capable of doing it before things get politically ugly is the actual open question. Right now, I’d say the evidence is mixed at best.
And there is a real, non-zero risk of the bad version where displacement outruns the ability of any government to respond, instability follows, and things get genuinely difficult for a stretch of years. I don’t think that’s the most likely outcome, but pretending the probability is zero would be dishonest.
What the Indian Workers’ Story Is Really Telling Us
Go back to where we started. Every individual worker training those robots made a perfectly rational choice. The money is real today. If they refuse, someone else takes the job, and the training happens anyway. There’s no individual incentive to opt out.
But stack a few million of those individually rational choices on top of each other, and you get a collective outcome that’s much harder to feel good about. That’s the pattern to watch for: not any single decision, but what happens when millions of reasonable, self-interested choices add up to something nobody actually chose.
That’s the real story underneath the headline. Not “robots are coming for jobs.” We already knew that. The real story is that the people closest to the displacement often have the least power to do anything about it except participate in their own replacement, because the alternative is having no income today while waiting for a future that may or may not show up on schedule.
Where That Leaves Us
I don’t have a tidy ending for this, and I’m suspicious of anyone who claims they do. What I do believe is that the conversation needs to shift. Less “Are robots going to take jobs?” (they already are) and more “who owns the robots, and what happens to the people whose work they replace?”
That’s a harder conversation. It’s also the only one that actually matters.
If you’re watching this unfold the way I am, with more questions than answers, I’d genuinely like to hear how you’re thinking about it. This is one of those moments where staying curious and a little uncomfortable is probably the right response, and certainty in either direction should make you suspicious of whoever’s offering it.
