Instead of controversial wealth or broad income taxes, a more politically viable solution for AI-driven job displacement is to levy a higher corporate tax rate specifically on companies whose profit margins surge after replacing workers with AI.

The core argument for a token tax is not to penalize AI, but to ensure the tax system doesn't artificially favor automation. It shifts the tax base from human labor (payroll, income taxes) to AI's productive capacity, measured in tokens, to prevent tax-incentivized job displacement.

Current tax structures penalize human labor but not machine labor, creating an incentive for automation. A tax on AI compute (tokens) would level the playing field, fund social programs for displaced workers, and is presented as a politically feasible bipartisan solution.

Governments rely heavily on taxing workers. As AI displaces jobs or suppresses wages, this primary revenue stream shrinks. AI creates value in areas like capital and corporate profits, which are taxed less heavily, leading to a potential government funding crisis even as the economy grows.

Taxing AI usage via a "token tax" is a flawed policy. It disproportionately harms the most ambitious and productive firms—those using AI to augment their human workforce and boost competitiveness. This creates a perverse incentive to avoid the very AI adoption that strengthens the economy.

Bill Gates advocates for a fundamental shift in tax policy to prepare for AI-driven job displacement. Instead of traditional income tax, he suggests taxing AI compute usage, or "tokens." This would create a new revenue stream to fund a social safety net for the millions of jobs he predicts will vanish.

The AI industry and the US government both require trillions in funding. This creates a paradox: the more successful AI becomes, the more it erodes the white-collar tax base by automating jobs, forcing the Treasury to borrow even more and intensifying the competition for scarce capital.

Emad Mostaque argues that the math for a tax-funded Universal Basic Income (UBI) doesn't work. Providing even a poverty-level UBI in the U.S. would cost $5 trillion, the entire federal tax base. Corporate taxes from AI giants wouldn't come close, necessitating a fundamental rethinking of how money is created and distributed.

Taxing AI tokens is a poor strategy as their cost approaches zero. A more sustainable model is to tax externalities created by AI, like a per-mile tax on autonomous vehicles, or to tax the enormous excess profits that AI will concentrate in a few companies.