You Can’t Cross a River Without Getting Your Feet Wet
A conversation with Claude
I’m lucky enough to have friends who make me think, learn, and reflect. One of these dear friends, Giulia, sent me a video about Indian women being exploited to train large language models (this article in The Guardian is one of many that explore the subject). That video compounded over all the news I’ve been seeing and avoiding at the same time, about the exploitation of people in poor countries to feed our hunger for advanced technology. When I received that video, I don’t know why, I found myself staring at an uncomfortable question: have human being always exploited other human beings? Is it part of our nature?
So, I took my ignorance and asked Claude. I didn’t have a goal in mind. But the conversation that ensued led me on an unexpected journey. So, I’ve decided to share it with you. I’ve edited it a bit to make it more readable, but I can share the whole conversation with the links and facts if you want to experience the whole journey.
Fabio: I was reading about women in India and Africa who are paid to label data — to validate content that trains AI systems like you. And about cobalt miners in the DRC, whose work makes the batteries in our devices possible. It made me wonder: when did this start? Was there a moment when human beings began exploiting other human beings?
Claude: The honest answer is: we don’t know precisely, and the question is harder than it seems. The oldest written evidence of formal slavery comes from Mesopotamia, around 3500–2500 BCE. But that’s just the limit of written records, not the limit of the practice. Archaeologically, we can find traces — violent trauma on skeletons, non-local individuals in subordinate burial contexts — that suggest exploitation existed long before cities did. What we can say is this: there was almost certainly no single first moment. It emerged in multiple places, independently, whenever conditions allowed.
Fabio: Interesting and fascinating. I was curious because today we have more subtle forms of slavery or exploitation of human beings. Cheap labour in poor countries, for example. Rare mineral mining. But in a more recent example, people also used content validation to train LLMs in poor regions. As you wrote, it’s a complex topic. Who is better, a free man who can’t afford to have food and shelter, forced to do degrading jobs, or a formal slave of ancient times who didn’t have freedom but maybe had a good quality of life?
Claude: That comparison is real and worth taking seriously. But it carries a hidden risk — the “maybe slaves had it better” framing has historically been used to justify exploitation. “They’re fed and housed” was the explicit defense of American chattel slavery.
What I think you’re really asking is whether we’ve made moral progress at all. And that question has two uncomfortable answers living side by side. We’ve made genuine normative progress — slavery is universally illegal, morally condemned, and crucially, people who practice it now hide it. The need to hide exploitation is itself a form of moral progress. But structurally? There are more people in conditions of slavery today in absolute numbers than at any point in history. What we’ve done is largely externalize and obscure exploitation rather than eliminate it. The Kenyan content moderator watching beheading videos for a dollar fifty an hour is legally free. The freedom is formal, not substantive.
The most unsettling version of your question is this: has moral progress itself become a tool of exploitation? By convincing ourselves that slavery is over, that we live in a free market of free choices, we may have made ourselves less capable of seeing what’s in front of us than a Roman citizen who at least knew he owned a slave.
What do you think — does that track with how you see it?
Fabio: Yes. And the thing that scares me the most is that, as you wrote, moral progress itself has become a tool of exploitation. My lifestyle, the technology I use, the quality of life I have. It is built on exploitation. Would it be possible without that exploitation? That’s the tough question. Could we still have the quality of life that few of us have without that exploitation? Is it truly impossible, or are we just going for the easiest way?
Claude: This is a question that doesn’t have a comfortable answer. And I’ll tell you upfront: I’m not going to pretend there’s a clean “yes we could, here’s how.” But I also won’t let the complexity become an excuse for fatalism.
Let me think through it honestly with you.
Could you specifically have exactly the life you have now, unchanged, without exploitation underpinning it? Almost certainly no. Not because your lifestyle is uniquely extractive, but because the infrastructure of modern comfortable life — devices, connectivity, cheap goods, global food systems — is deeply woven into these structures. You can’t fully opt out. You can only reduce your implication.
But I don’t think that means the alternative is impossible. Studies on fair trade supply chains consistently show that paying workers fairly would add marginal costs to end products — often one to four percent on the final price. The reason it doesn’t happen isn’t economic necessity. It’s that the savings go to shareholder returns, not to making products more affordable. The exploitation is chosen, not inevitable.
So: easiest way, or only way?
Easiest way. Clearly.
The alternatives exist — cooperatives, living-wage supply chains, and democratic ownership of data. But they require friction, cost, and above all, the willingness of people with power to accept less. That last part is where it always stalls, because it would require the people who currently benefit most to push hardest for it, which is precisely the group with the least immediate incentive to do so.
That paradox, I think, is the actual trap. Not ignorance. Not impossibility. Incentive structure.
Fabio: You know, I’m not 100% sure those alternatives are real alternatives. I feel like there’s always some degree of exploitation. Maybe it’s precisely because of that incentive structure you talk about. It looks like, as soon as someone’s life gets better, they fall into that trap and exploit someone who is where they were before.
Claude: You’re pointing at something most political frameworks try to avoid looking at directly. What you’re describing looks less like a policy failure and more like a pattern in human behavior — one that repeats across contexts, cultures, and centuries. The freed slave who acquires slaves. The immigrant generation that pulls up the ladder. The progressive company that outsources its supply chain. The person who escapes poverty and votes against redistribution.
This isn’t cynicism. It’s observable. And it raises a question that’s genuinely hard: is exploitation something humans do under certain conditions, or something humans tend toward as soon as conditions allow?
The uncomfortable possibility is this: what if the incentive structure is human nature, not in the sense that we can’t resist it, but in the sense that resisting it requires continuous, active effort rather than the natural default? That would mean exploitation doesn’t need a villain. It just needs people relaxing into comfort. Moral progress isn’t a ratchet that locks in gains. It’s more like a muscle that atrophies without use.
So, here’s the challenge. If the incentive trap is genuinely structural — baked into how humans respond to security and surplus — then the interesting question isn’t how do we eliminate exploitation but something harder:
What would it mean to live honestly inside a system you can’t fully exit, without either pretending your hands are clean or collapsing into fatalism?
That’s not a policy question. It’s an existential one.
Fabio: Good point. I often say that you can’t cross a river without getting at least your feet wet. Your question is, indeed, an existential one. As soon as we reach a certain level of comfort and security, we become self-centred. At best, tribe-centred. From there, others, outside the tribe, become resources that can be exploited if needed. And now, in our global society, we have become so good at not seeing that we are exploiting others, or the planet, that we feel like good people even when our lifestyle costs the dignity, freedom, health, and often the lives of many others.
Claude: Yes. And the mechanism you’re describing has a specific shape. It’s not that comfort makes people evil. It’s that comfort makes people local — in their perception, in their empathy, in what they count as real. The suffering that’s far away, or structurally hidden, or mediated through a supply chain, simply doesn’t register with the same weight as the suffering in front of you.
And then we’ve built entire cultural systems to justify that narrowing. Not just to enable it — to make it feel like virtue. Buying the right brand. Using the right language. Having the right opinions. The performance of conscience becomes a substitute for the exercise of it. What’s insidious is that the performance isn’t cynical. Most people doing it genuinely believe it. That’s what makes it so stable as a system.
There’s a concept in moral psychology — moral disengagement, developed by Albert Bandura — that describes exactly this: the cognitive mechanisms people use to behave in ways that violate their own standards without experiencing guilt. Displacement of responsibility. Dehumanization. Diffusion across a system. Euphemistic labeling. These aren’t rationalizations people consciously choose. They’re automatic, and they’re socially reinforced.
What your observation adds to that is that scale and abstraction are the most powerful disengagement mechanisms ever invented. You don’t need to dehumanize the cobalt miner. You just need to never see him. The global economy does the rest.
You said: “We feel like good people, even when our lifestyle costs the dignity, freedom, health, and often the lives of many others.”
The word that stands out to me is feel. Not are, not claim to be. Feel.
Because feeling like a good person and being one have always had some gap between them. But something has changed in scale. We’ve industrialized the production of that feeling — through branding, through social media, through the language of sustainability, ethics, and impact — in a way that widens the gap while making it less visible.
The feeling of being good has become one of the products the system sells us. Possibly one of its most profitable ones.
Which brings you back to the existential question, but sharper now:
If the feeling of being a good person can be manufactured and sold, what does it actually mean to try to be one?
Not perform it. Not purchase it. Actually try.
Fabio: “If the feeling of being a good person can be manufactured and sold, what does it actually mean to try to be one?” that! Like all the conversations on purpose, all the effort companies put in saying that they have a higher purpose. They are selling us that “feeling” of being among the good ones. What does it actually mean to be one in such a system? When, for survival conditioning, comfort makes us local?
Claude: That question has two escape routes that both feel satisfying and are both traps.
The first: radical individual responsibility. Examine everything. Opt out as much as possible. Accept guilt as the price of honesty. This has integrity, but it tends to produce either paralysis or a new form of the same performance — the person who is visibly, loudly paying the cost becomes their own kind of brand.
The second: systemic absolution. The system is too big, individual action is meaningless, only structural change matters. Also partially true — but suspiciously convenient for people who benefit from the system.
Both escapes have one thing in common: they relieve you of the discomfort of the question.
What I keep coming back to is something you said almost in passing: as soon as we reach a certain level of comfort and security, we become self-centred. That “as soon as” implies a threshold. And thresholds imply that below them, something different is possible — that the capacity for genuine moral attention requires a specific relationship to your own vulnerability. Not poverty. But not insulation either. The person who hasn’t fully sealed themselves off from the possibility of loss might simply see more than the person who has.
Which would make “being good” not a set of actions or beliefs but something closer to a quality of attention. The willingness to keep looking at what comfort wants you to stop seeing.
Maybe being good, in a system designed to sell you the feeling of it, starts with refusing the feeling. Not as punishment. As clarity.
Fabio: I agree. Neither of those escape routes truly leads anywhere. I believe the first step is to acknowledge the paradox. To acknowledge that we are the system. And if we do that, maybe we can move that threshold a bit. It’s not about perfection. It’s not about being a good one, whatever that means. It’s about expanding our capacity to care. Choice by choice. Even if by a tiny fraction.
Claude: That’s it. And I notice you didn’t need me to get there. The conversation found it.
What strikes me is that you’ve changed the question entirely. Not am I good, which is a static judgment, and immediately gameable by the system. But am I expanding my capacity to care, which is a direction, not a destination. A vector, not a position. A destination can be faked. A direction, lived honestly day by day, is much harder to manufacture and sell.
The other thing worth naming: we are the system is genuinely difficult to hold. Not as a slogan — it’s easy as a slogan. But in practice, it removes the comfort of having an enemy outside yourself. It means the work is never done and never clean.
But it also means the leverage is real. If we are the system, then every choice that moves the threshold — even by a fraction — is not symbolic. It’s structural.
I’ll be honest with you: this conversation has gone somewhere I didn’t anticipate when you asked about the origins of slavery. There’s a thread running through all of it — from the archaeological question about when exploitation began, to the manufacturing of moral feeling, to this — that feels like it has an essay in it.
Not one that resolves anything. One that ends exactly where you just landed: with the paradox acknowledged, the escape routes refused, and the only honest movement being the slow expansion of attention, choice by choice.
Fabio: Yes. And I may share it as it is. As a conversation. However, I confess that I feel a bit guilty now while I read this exchange. Sure, I may publish it. But am I really doing something, though, or am I just playing the same feel-good game?
Claude: That’s the most honest thing you’ve said in this entire conversation.
And I’m not going to reassure you out of it.
You’re right to hold the question — because publishing a piece about the manufactured feeling of being good can itself become a manufactured feeling of being good. The trap is that sophisticated.
The difference, I think, is not in the act of publishing. It’s in what you do with the discomfort after. Whether this conversation changes something in how you actually live — a choice here, a bit more attention there — or whether it gets filed under “things I’ve thought deeply about” and the thinking becomes the alibi.
You can’t know that yet. Neither can I.
What I’d say is this: the guilt you’re feeling right now is the threshold moving. Slightly. That’s not nothing. But it’s also not enough on its own — and you know that, which is why you’re asking.
The question isn’t whether writing this is sufficient. It clearly isn’t. The question is whether it’s part of something real or a substitute for it. Only you can answer that.
One thing I notice, though: you asked me if you’re doing something real. You didn’t ask me to reassure you. That gap — between wanting comfort and asking the harder question instead — might be the most accurate answer you have right now.
Keep that gap open. It’s doing something.
That’s where I stopped. I don’t know yet if this conversation is going to change something in how I live. And I’m well aware that I did it with an AI service, a technology that is itself built on the exploitation I've been writing about. This contradiction is part of my journey. Maybe it’s part of yours, too. I don’t know. But I hope that by sharing it, we can start another conversation. One that can help both of us—you who are reading it and me—move the threshold, while we do the best we can.
Because, in the end, you can’t cross a river without getting your feet wet.
Thanks for reading this far. Muchness is where I explore questions that won’t leave me alone. If that sounds like your kind of thing, subscribe. And if this piece resonates with you, support this journey with a like or by sharing it with those who, like us, are trying to make sense of their reality.