Tech luminary Bill Gates has become the latest proponent of an AI tax that could, in theory, protect the labor market. But is a token or excise tax really the best way to contain the spread of AI? Should the U.S. even try to contain it?
Gates’ proposed AI tax would counteract a system that, in theory, expedites the redundancy of human labor. Because companies can write off business expenses (like AI), "The tax system nudges you toward replacing people with machines," Gates says.
And while companies have to pay payroll taxes on human employees, AI carries no such fiscal obligations.
Gates is certainly not the first one to propose an AI tax.
Congress has suggested various levies, like the “AI Tax and Work Protection Act”, which would apply a tax on either AI "tokens" (units of data/computing power processed or sold) or total AI revenue, and use the proceeds to fund job creation programs for workers displaced by AI.
Sen. Bernie Sanders (I-VT) has proposed a one-time 50% tax on the stock (equity) of firms with over $200 million in annual AI gross receipts.
But is a comprehensive AI tax really in America’s best interests? The Cato Institute points out that AI taxes could be fundamentally counterproductive. First off, because taxing intermediate goods “raises costs before any final good or service is produced,” which then filters through downstream products.
Moreover, levies of this sort would be capital taxes, and historically, these fees are passed on to workers through lower investment and lower wages. And even if we taxed 100% of the expected $1.4 trillion in AI spending next year, the blowback could be intolerable. Tax rates like that could “dramatically shrink projected AI spending and, thus, expected revenue. AI development wouldn’t stop; it would just shift offshore.”
Maybe that’s the whole point – critics want to impede AI’s growth. They certainly want to protect the labor market, a noble enough goal. But taxing it and deliberately hindering it could expedite the problem.
AI is here. There’s no getting around it. I probably don’t need to remind you that PSD employs an OpenAI-powered chatbot, “PowerPal,” to improve the user experience. Most of our readers work make expeditious use of AI in their daily workflow. Much of the research for this very article was conducted via search engines optimized by AI.
And if arbitrarily limiting AI’s growth just cedes the advantage to geopolitical rivals and kills more jobs than it saves, then it might be the wrong move. I’m as big an AI critic as any, but perhaps the goal should be ensuring its responsible use, rather than hindering it altogether.
