It's 8:47 AM in Electronic City, Bengaluru. Infosys buses have already unloaded their first wave. Thousands of people cross the turnstiles with credentials hanging from their necks, laptops on their shoulders, and coffee in hand. Analysts, developers, testers, and project coordinators. The campus has its own rhythm, its own grammar. Someone reviews a ticket in Jira. Another joins a call with a client in Frankfurt. At TCS Whitefield, a few kilometers away, the scene repeats with minor variations. This isn't Silicon Valley. Nor is it the collapse some had predicted.
That same morning, in San Francisco, someone writes an essay about universal basic income for those workers. The workers didn't read the essay.
This raises an uncomfortable question. Who is the universal basic income debate really addressing when it's presented as a response to technological displacement? Is it a policy proposal with concrete recipients, or a conversation that certain actors have with themselves, about people who weren't invited and in a language those people didn't choose?
May 2026 data complicates the dominant narrative. NASSCOM reports 5.8 million jobs in India's tech sector for fiscal year 2025, with a net gain of 126,000 positions. IBPAP projects growth from 1.4 million to 2.5 million jobs in the Philippines by 2028. These are aggregates, not anecdotes. The story of emerging economies gutted by AI doesn't hold up against these numbers. What exists is internal reconfiguration: changes in profiles, tasks, and hierarchies. Not contraction. The difference matters, because policy arising from labor catastrophe differs completely from that arising from complex transition.
In Bengaluru, salaries range from $3,000 to $12,000 annually depending on level and company. There are restructurings and teams that shrink. When you ask for the official cause, AI rarely appears. They talk about operational efficiency, function consolidation, or strategic adjustment. Nobody proposes universal basic income in Karnataka. Nobody in those hallways expects a check from a government that discusses their future in English from the other side of the Pacific. The image recalls Sleep Dealer: work flows, connectivity flows, but citizenship in the debate doesn't.
Manila operates under different logic. IBPAP plans expansion, not containment. The government maintains the 4Ps program, a conditional cash transfer targeted at families in poverty. It's assistance against structural poverty, not compensation for future technological transformations. Nobody receives a check because ChatGPT exists.
Kenya is the most documented case and also the most misinterpreted. GiveDirectly has been operating for years, has reached 23,000 people and delivers between $22.50 and $40 monthly via M-Pesa. A joint UC Berkeley and Oxford study published in October 2025 found that each dollar transferred generates $2.50 in local economic activity. The finding is solid. Still, GiveDirectly is explicit: this is poverty relief, not compensation for technological displacement. Beneficiaries in Siaya County didn't lose jobs at call centers to automation. Many never had formal access to the labor market.
In the rich north, pilots continue their course. Stockton delivered $500 monthly to 125 families for two years. OpenResearch distributed $1,000 monthly to 3,000 people across multiple states. San Francisco has discussed local proposals. None scaled to federal policy. Altman and Musk frame them as anticipatory responses to AI, as buffers against massive transformations. The IMF estimates that nearly 40% of global employment has significant exposure. However, these pilots cover less than 5% of that workforce and concentrate in high-income countries. The proportion doesn't add up.
The real asymmetry isn't in the numbers but in the political grammar. The same instrument is discussed as three different things depending on geography: anticipatory compensation in the north, poverty relief in Kenya, and traditional social assistance in the Philippines. Three discourses that use similar words but don't dialogue with each other, that don't share diagnosis or recipients or coordination processes. Each actor uses the instrument to solve their own problem. The result is a global conversation that sounds coherent from the outside and is completely fragmented within.
Argentina is the case that most disrupts the map. Stargate Patagonia represents a $25 billion investment in artificial intelligence infrastructure. The country receives the data centers, the cabling, and the computational capacity. It doesn't appear, however, in any compensatory framework: not in Silicon Valley discourse, nor in IMF reports, nor in debates in Washington or Brussels. The current government reduces public spending and eliminates social programs. It absorbs the infrastructure of the next digital economy without being included in any conversation about how to distribute its benefits or mitigate its costs.
There are aspects of this map I don't have completely resolved. It's unclear whether the fragmentation is deliberate or simply the result of different actors solving different problems without coordination. It's probably both, and that ambiguity is part of the problem. Those who study ruins of ancient resource distribution structures notice something uncomfortable: current debates repeat old patterns about who gets left out of the conversation. Stones don't lie, but historians sometimes do.
The conversation takes place where fiscal capacity and political voice exist. Not where the changes occur.
