“Forget retirement: even money will lose its meaning, because AI and robots will drive costs to zero.” That roughly captures the recent debate sparked by Elon Musk’s remarks. At first glance, the claim is mesmerizing. But introduce a little economics, energy, and politics into the story and a very different set of questions emerges from the confident prediction.<\/p>\n\n\n\n
First, generative AI is not free. Training and operating models requires electricity, data centers, cooling systems, transmission infrastructure, chips, memory, water, and land. Even if labor becomes cheaper, the costs of production are merely shifted elsewhere, including in work that remains labor-intensive. And electricity, one of the system’s most important inputs, is not simply something with a price; it is also a scarce resource. The claim that marginal costs will fall to zero may be partly true for a narrow range of digital activities, but it cannot automatically be extended to the entire cost of living.<\/p>\n\n\n\n
Second, a society’s ability to sustain itself depends not simply on production capacity but on purchasing power. If large segments of the population lose their incomes, companies may be able to produce far more\u2014but to whom will they sell? That is where the promise of abundance begins to fracture. As earnings from labor weaken, the stability of the system comes to depend on constructing a new regime of distribution. Doing so requires painful political decisions that would transform the existing order at its foundations.<\/p>\n\n\n\n
Third, when someone proposes that the state support everyone through a universal citizen’s income, the question immediately becomes: where will the taxes come from? If taxes on labor decline, the tax base will have to shift toward profits, rents, consumption, and ownership. Taxing multinational firms is already difficult and may become harder as they operate on a more global scale. If the state’s capacity to collect taxes remains weak, AI-generated abundance could be abundance accompanied by even greater inequality. Some people may gain access to cheap services while vast populations are left without income.<\/p>\n\n\n\n
Fourth, there is the international division of labor. The centers of generative AI are concentrated in particular countries. Revenues from chips, cloud services, platforms, and intellectual property are likely to accumulate in these core economies. Countries on the periphery will use the technology as an imported intermediate input, paying subscriptions and licensing fees and transferring a substantial share of their productivity gains back to the center. Under those conditions, the formula “AI reduces costs, corporate profits rise, governments collect taxes, and citizens receive support” begins to break down.<\/p>\n\n\n\n
For the world’s richest person, it is easy to announce that your retirement savings and mine will become meaningless. But this is an empty prophecy unless we ask how abundance will be distributed. In a post-GenAI world, which taxes can be collected globally, and in which countries will the resulting value accumulate?<\/p>\n\n\n\n
The future will be shaped not by the physical disappearance of scarcity but by how scarcity is reproduced in different people’s lives. Citizens therefore need to examine how resources are distributed within their countries, the capacity of governments to raise revenue, and their countries’ positions within global value chains.<\/p>\n\n\n\n
At the present pace of change, that future may not be very far away.<\/p>\n\n\n\n