OpenAI’s Sam Altman has expressed his support of a Universal Basic Income (UBI) in the United States on many occasions. Here’s one of them.
I want to explain why he’s wrong and why I think that a UBI in the United States is highly unlikely for the foreseeable future. Here’s the TL;DR version. UBI only works if implemented globally, because any country that offers a meaningful unconditional income becomes a magnet for global migration. But global UBI is impossible to coordinate, impossible to fund, and impossible to enforce. Therefore, UBI is inherently impractical and cannot be implemented in the real world. Anything less than global implementation is not universal, not sustainable, and ultimately self-defeating.
The economic problems
The price tag for a UBI is staggering. Even for a minimal $10,000 per year program for 260 million adults it would cost $2.6 trillion per year. In the absence of real material production rising that would produce inflation. Indeed, increases in the prices of housing and other essentials diminish the value of the UBI. It becomes a positive feedback situation.
Furthermore, raising taxes to pay for it is not a practical solution. The topmost income earners (or possessors of wealth in the case of a wealth tax) can and do leave.
From a technical standpoint a value-added tax is the only one capable of generating enough revenue and not only has that been historically unpopular in the U. S. it would run into the same positive feedback dilemma mentioned above.
The resistance to cutting presently existing programs is enormous. We are seeing that right now. The costs of a UBI would exacerbate that.
The social problems
The United States is, perhaps, peculiarly ill-suited for a UBI. Here in the U. S. identity, dignity, self-worth, and status are closely tied to work and the income realized from work. A software engineer who loses a job paying $180,000 a year to AI will have emotional and social difficulty receiving the same “basic income” as an individual who never graduated from high school or, possibly, worked a paying job in their lives.
The political problems
There is little or no coalition to support a UBI in existence. At present the overlap among progressives, moderates, and conservatives is so small it’s hard to imagine such a coalition forming. There are already substantial, active constituencies for Social Security, Medicare, and the Affordable Care Act. There aren’t any for a UBI.
The irony of LLM AI
The irony of LLM AI is simple to state. It’s expensive. The data centers required for LLM AI:
- produce huge, concentrated loads
- require constant cooling
- draw power even when human usage drops
- strain transmission lines
- require substation expansions and new peaker plants
Their demands will tend to raise the price of electricity. LLM AI threatens entry-level white collar jobs and the incomes associated with them, popularizing the idea of a UBI among progressives, tech-utopians, displaced workers, and futurists.
It should also be mentioned that a UBI will induce migration from areas of the United States where the climate is harsher, e.g. North Dakota, Minnesota, to places where it is milder, e.g. California, Florida, and North Carolina producing price increases in the destination areas and fiscal collapse in the origin areas. Furthermore, those destination areas are those most vulnerable to increases in the price of electricity and vulnerable to outages.
AI raises the floor of human needs while lowering the ceiling of what society can afford.
In conclusion it should be obvious why I think that a Universal Basic Income is unlikely, at least in my lifetime. More likely will be reinforcement of existing programs, however poorly constructed they might be.
