WASHINGTON, DC - JULY 23: U.S. President Donald Trump speaks during the "Winning the AI Race" summit hosted by All‑In Podcast and Hill & Valley Forum at the Andrew W. Mellon Auditorium on July 23, 2025 in Washington, DC. Trump signed executive orders related to his Artificial Intelligence Action Plan during the event. (Photo by Chip Somodevilla/Getty Images)

Getty Images

Capitalism is too fast for governments. Which means Artificial Intelligence (AI) is way too fast.

As Nvidia CEO Jensen Huang routinely points out, thirty days in AI is an eternity. It’s his way of conveying the crucial truth to Nvidia employees that in their industry, stasis is the rapid path to obsolescence.

The speed at which AI is evolving speaks to why the federal government should have no investor or ownership role in AI. This rates stress as President Trump and Sen. Bernie Sanders call for just that. No thanks.

It’s plainly lost on Sanders and Trump that if the federal government were given an ownership role, those in its employ wouldn’t have a faint clue what to own as is. See Jensen Huang again.

The federal government’s vision of AI is so dated as to be primitive. It recalls a recent comment from Nader Khalil, co-founder of Brev.dev, a Nvidia portfolio company. He observed that more lines of code have been written this year than in all the history of computing. It vivifies what Huang means about the meaning of time in the AI sector. The present is the past.

Which means government investments in AI would be dated, in addition to inept. Precious capital propping up the past.

To which Sanders and Trump might say that federal government would sign up all manner of experts in the AI sector to ensure an elite collection of AI properties for taxpayers to own. The view isn’t serious.

For one, no AI investor of any substance would be in the federal government’s employ as is. To do so would be to potentially forego billions. Only for the story to get worse.

As is nearly always the case with politicians, they only see successes. Trump and Sanders see OpenAI, Anthropic, and Cerebras to name but three AI high-flyers, and they tell themselves government should have a piece of this surely lucrative action. AI investing made easy…

Except that as the investors in all three could tell them, those are the seen. The unseen is all the investments in high-potential technology companies that don’t just fail, but fail quickly.

Which is where federal ownership of AI investments becomes horrific. Exactly because most venture investments in the technology space fail, government cannot be an owner. It can’t simply because the relentless failure in the technology sector is the sector’s sustenance. How else to produce abundant wealth (think information) absent routine mistakes?

Except that government doesn’t do mistakes. Instead, it subsidizes failure repeatedly. As opposed to creating wealth, politicians want to create jobs. It signals even more precious capital stuck in the past, as opposed to flowing to what’s next.

The simple, crucial truth lost on politicians terrorized by company failure and unemployment is that what we call an “economy” gains essential strength from periods of weakness. How else to free up human and physical capital for direction to its best use? Except that politics and periodic weakness don’t mix.

Which means politics and AI don’t mix. As every great commercial leap throughout history reminds us, there’s a lot of failure on the way to success. Since politicians can’t stomach the failure part, their investments will cruelly and dangerously restrain the successes.