AI companies compete with the US Treasury for capital, driving up interest rates the government can't afford. Simultaneously, AI aims to eliminate white-collar jobs that form the core of the federal tax base. This creates a "snake-eating-its-own-tail" dynamic that pushes the US closer to a fiscal crisis.

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.

As companies replace human workers with AI 'robots,' they eliminate a crucial source of government funding: payroll taxes. This trend threatens the solvency of programs like Social Security, which rely on a large base of human workers to support a growing retiree population.

By taxing wealth (e.g., capital gains) at a lower rate than labor (e.g., income), the US tax system creates a "thumb on the scale" that subsidizes automation. This policy actively encourages companies to replace workers, exacerbating job displacement and inequality.

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.

Despite clear warnings about AI's potential to erode tax bases, policymakers are unlikely to act proactively. The prevailing hope is that AI-driven economic growth will solve fiscal problems. Meaningful, structural tax reform will probably only occur reactively, after a government faces a severe funding crisis.

Following Amazon's model, AI-native companies will reinvest all available cash into acquiring more compute power for a competitive edge. They will operate in a perpetual land-grab mode and never need to show a profit, making them impossible to tax effectively and rendering corporate taxation an obsolete funding mechanism for the state.

Economist Ben Harris warns that AI-driven growth may disproportionately benefit owners of capital rather than labor. Because capital income is taxed at a much lower marginal rate than labor income, this shift in the composition of national income would lead to lower-than-expected tax revenues, even amidst strong overall economic growth.