In 1927 Fritz Lang captured the future with a silent camera. In Metropolis workers descend into the depths to feed the Machine while industrialists' children stroll through hanging gardens. When the masses begin to stir, the elite doesn't improve their living conditions. It builds a false Maria, a human-looking robot that preaches patience and promises that change will come. The phrase that closes the film sounds beautiful: the mediator between the head and the hands must be the heart. It takes just a second of reflection to notice that the one speaking these words is exactly the one who designed the machine that devours men.
That's what UBI represents today. Not the idea itself, but the precise moment and the voice from which it's being proclaimed.
The conversation about universal basic income stopped being marginal between 2023 and 2026. Goldman Sachs projected in March 2023 that three hundred million jobs face automation by artificial intelligence. The International Monetary Fund raised the figure in January 2024: sixty percent in advanced economies, forty percent in emerging ones. The World Economic Forum estimated ninety-two million jobs displaced by 2030. Concrete announcements piled up: IBM eliminated seven thousand eight hundred positions in 2023; Klarna replaced the equivalent of seven hundred people's work with a chatbot in 2024; Duolingo cut ten percent of its translators that same year; TCS India announced twelve thousand fewer employees in 2025 and Oracle India, ten percent of its workforce. This noise didn't sprout from an academic debate that matured in silence. It's the political response someone prepared for a structural crisis everyone sees coming.
The question that matters now isn't whether UBI works. That comes later. What matters is who's speaking and from where.
Elon Musk declared it inevitable from Dubai in 2017. Sam Altman links it to Worldcoin, his biometric identity project conceived as distribution infrastructure, and posted in January 2025: We are now confident we know how to build AGI. Andrew Yang brought it to his presidential campaign in 2020. Yanis Varoufakis insists on distinguishing classic UBI from Universal Basic Dividend because he understands that the money's source defines the proposal more than the check's amount. There's a constant: they speak from San Francisco, Athens, Manhattan or Dubai. The worker from Manchester doesn't appear in the conversation, not because the problem doesn't affect him, but because it affects him so much he barely has time to write about the future of work.
From Mexico these dynamics look different. Technological promises arrive wrapped in salvation language while real costs are distributed silently and unequally. I recognize this pattern in other contexts: when a solution presents itself before the problem is completely visible, it's worth asking who benefits from that being the only alternative on the table. Not because the solution is necessarily bad, but because other possibilities disappear before being evaluated.
UBI isn't a Silicon Valley invention. Thomas Paine proposed it in Agrarian Justice in 1797. Milton Friedman formalized it as negative income tax in 1962, explicitly because it was compatible with a free society's philosophy. Martin Luther King defended it in 1967. There's a long genealogy with very different motivations. Almost no one mentions, however, the parallel post-monetary line. Jacque Fresco, with his Venus Project since the seventies, proposed eliminating money completely and letting computers allocate resources. He wrote that it's doubtful that in the latter part of the twentieth century people would play any significant role in decisions. Two responses to the same question: one universalizes money, the other suppresses it but hands decisions over to machines. This series will compare both with a third that rejects both premises.
Before getting there, there's a syllogism nobody in Davos puts on the table. The public promise runs like this: artificial intelligence will eliminate jobs, therefore a check is needed to replace lost income. Sounds logical. What's omitted is everything it costs to build the scenario where that check becomes necessary. OpenAI's Stargate project with SoftBank, Oracle and MGX, announced from the White House in January 2025, committed five hundred billion dollars just for infrastructure within the United States, approximately Argentina's annual gross domestic product. NVIDIA adds another hundred billion in chips. AMD allocates six gigawatts in GPUs. Microsoft, Meta and Google contribute comparable figures separately. The International Energy Agency projects that data centers will consume nine hundred forty-five terawatt hours by 2030, more than Japan's annual electricity. In Loudoun County, Virginia, those centers already use twenty-one percent of local electricity, more than all combined residential consumption. Residential bills will rise eighteen dollars per month in Maryland and sixteen in Ohio. A Carnegie Mellon study anticipates an eight percent increase in the United States by 2030 that will exceed twenty-five percent in Virginia.
The people whose jobs will be displaced are already paying, through electricity bills that rise month after month, for the infrastructure that will displace them. Then they'll be offered a UBI financed with public debt or taxes. Once with the electricity, again with taxes. And that's only if they live in rich countries. The supply chain that sustains all this—cobalt from Congo, lithium from Uyuni salt flats, rare earths from Inner Mongolia, water from Arizona—crosses territories that won't participate in the check. Stargate's centers have already begun installing in Argentine Patagonia with twenty-five billion in investment. The question nobody asks at conferences is how many liters of Arizona water equal how many pesos in a Kansas City typist's monthly check.
Pixar drew it in 2008. In WALL-E a single corporation provides everything to humans sitting in floating chairs who no longer walk, think, or decide. The artificial intelligence called AUTO obeys a hidden directive for their own good. In 2026 that stopped being a comfortable metaphor. The conversation about UBI converges exactly toward that model: the same corporations whose infrastructure displaces work will be the ones distributing the check. Worldcoin proposes biometric identity as the access key. The question the film left unanswered—who decided what's good for the humans on the Axiom ship—is precisely the one the debate elegantly avoids.
This is the first installment of a series. The following texts will dismantle the question layer by layer: the pilots conducted in Finland, Stockton, Kenya and in Texas and Illinois studies, along with what the data actually shows. The three geographies where UBI means radically different things. The human dimension of prolonged unemployment, where the Marienthal study in 1933 already anticipated what Pixar visualized decades later. The opposing visions of AGI according to Yann LeCun and according to Sam Altman, and why it matters who's right. The three proposals compared and why they're not simple variants of the same thing. And at the end, what none resolves: the political question the technical debate avoids.
For now I maintain Lang's image: the false Maria preaching patience while the Machine keeps running. And a question I still haven't resolved: is there any version of UBI that isn't manufactured by whoever built the machine, or does the question itself already contain the trap?
