Key Takeaways

  • Bridgewater Associates Managing Chief Investment Officer Greg Jensen predicts that 14% of current jobs will face radical disruption within three years.
  • The existing corporate tax code subsidizes automation: businesses pay payroll taxes on human staff, while AI compute incurs no labor taxes.
  • Jensen compares our current societal blind spot to February 2020 before COVID-19 lockdowns, warning that white-collar displacement will hit faster than previous trade shocks.
  • To prevent unchecked wealth concentration, Jensen proposes a token tax on AI compute to level the playing field between human workers and autonomous software.

The Tax Code's War on Human Labor

When you hire an engineer or an analyst, you pay a penalty. You cover employer payroll taxes, healthcare contributions, and local employment fees. When you buy API credits or rent server clusters to run autonomous agents, the tax code treats those costs as ordinary operational expenses.

Greg Jensen points directly at this asymmetry. “One thing that's obvious is you shouldn't be putting human labor at a disadvantage to machine labor,” Jensen says. “And it is today. We tax human labor, that's a disincentive. Whatever you tax, you're disincentivizing.”

By taxing payroll while leaving compute untaxed, governments create an artificial incentive to replace people with models. Jensen calls this imbalance out plainly: “You tax labor, human labor, and not machine labor, you are disincentivizing one versus the other. Why are we doing that?” If software functions as an employee, it should face an equivalent fiscal burden. “If a company is hiring it as a worker, it should at least pay taxes proportionate to income taxes that humans pay, or else you're prioritizing machine labor over human labor.”

The WTO Shock at Compute Speed

Jensen compares the incoming wave of white-collar displacement to global trade shocks from recent decades. “I believe in 3 years 14% of current jobs will be radically changed,” Jensen notes. “We just experienced this with the WTO. If China coming on has just like a source of labor, you create a new source of labor, it's disruptive.”

When China entered the World Trade Organization, millions of manufacturing jobs moved overseas over the course of two decades. That transition triggered deep political and economic fractures. AI creates a vast supply of white-collar labor, but the displacement cycle will unfold in months rather than decades. Jensen warns that society's current lack of urgency mirrors the complacency seen in February 2020, right before COVID-19 shut down the global economy.

The fix is not banning AI, but taxing its usage to cushion the societal blow. “AI is taking these jobs and we're not even taxing the AI,” Jensen argues. “Like that feels like a crazy path to be on, and we can fix that path.” A token tax creates two immediate benefits: it slows down premature replacement of humans where machines are only marginally cheaper, and it generates the revenue needed to fund worker transition programs.

What to Do With This

Audit your five-year headcount and technology budget today. Calculate your total cost per human employee (salary plus 15% to 25% in payroll taxes and benefits) versus your estimated API token costs. Then re-run that model under the assumption that governments will introduce a 10% to 20% excise tax on commercial AI compute before 2028. If your startup's margins only work because compute is completely tax-advantaged over humans, your cost model is vulnerable to incoming fiscal policy.