Get Paid to Use AI Agents? Why Nvidia CEO Jensen Huang Just Pitched a Game-Changing New Compensation Model for Engineers.

Tech workers have always been paid in two currencies: salary and stock options. Now, there’s a third currency entering the chat: compute.

Last week, Nvidia (NVDA) CEO Jensen Huang floated a new compensation idea that could fundamentally change how white-collar workers get paid. Instead of just earning a monthly paycheck, Huang’s idea is to give staff an annual budget of artificial intelligence (AI) tokens. That means Nvidia would essentially be giving its employees prepaid access to the world’s most powerful AI agents and systems.

Huang didn’t mince his words about how much AI firepower he’d hand over to his workers, either. Speaking at Nvidia’s latest GPU Technology Conference, the AI boss said he wanted to start giving his engineers half their current base pay in AI tokens on top of their existing salary.

Translation: If you’re a Big Tech engineer making a base salary of $400,000, Huang thinks you should be getting paid another $200,000 worth of AI firepower.

At first glance, this looks a lot like a daft Silicon Valley gimmick. But dig a little deeper, and Huang’s idea actually says a lot about evolving competition and what productivity should (or will) look like in the age of AI. 

To a lot of white-collar workers, this is going to seem a little scary. But it’s a future worth taking a closer look at.

Why Is Nvidia Pushing This Now?

If you watch Nvidia closely, you’ll know that Huang isn’t the kind of guy who makes off-the-cuff remarks. If he’s talking HR in front of an audience, it’s not some half-baked, quasi-futuristic idea. It’s cold, hard strategy.

At its core, Nvidia is all about turning AI into infrastructure. If the company is ultimately successful, logic dictates that access to compute will become just as important to workers as access to capital. That’s where his new token model comes in.

Instead of scaling teams by hiring more engineers and paying bigger salaries, Huang thinks the way forward is to give workers AI agents and let them start automating. That multiplies output without multiplying headcount — meaning that one Nvidia engineer could easily do the work of 10 people.

Huang is positioning this strategic shift as a win-win. Engineers with more compute can amplify their output by leaving all the repetitive stuff to AI. They then get to focus on high-value and more intellectually challenging tasks. A lot of engineers are going to love this because it gives them more career leverage.

But you’ve probably already spotted the catch: If one engineer and his AI tokens can do the work of 10 workers, 90% of your workforce becomes redundant. That’s the part of this conversation that’s getting louder on Wall Street.

After all, if Nvidia’s new concept for a hybrid digital workforce gains traction, the economics that underpin human labor are going to start to change very quickly. And everybody knows Nvidia is a trendsetter, right?

How Does Nvidia’s New Idea Fit Into The Wider Labor Market?

Huang might be leading the charge into this brave new world, but Nvidia’s new compensation model isn’t happening in a vacuum. 

Over the past 12 months, there has been a flurry of warnings about how AI advances are impacting the job market. Researchers at Goldman Sachs (GS) have warned that AI is currently on track to automate a quarter of all work hours in the U.S. With 300 million jobs exposed to automation, the investment bank has said to expect between 6% and 7% of those 300 million jobs to disappear over the next few years.

Layer Huang’s new AI token scheme on top of that, and you get a very different labor market. Corporations will be measuring output by token rather than employee, and so access to AI will become both a competitive advantage and a requirement for staying employed.

It also impacts what it actually means to work. Instead of doing work directly, the respective roles of white-collar workers who are lucky enough to stay in the game will become more like orchestrators and agent managers than engineers. That sounds really efficient, but it raises deeper questions around what happens to workers who are unable (or unwilling) to adapt.

We don’t have those answers right now, but that’s not a gradual shift. It’s a step change, and it’s something both employers and staff are going to have to figure out really quickly.

What This Means For Wall Street

It’s one thing to argue about the hypothetical evolution of work as we know it, but it’s also important to zoom out for a bit of context. Huang’s AI compensation proposal isn’t just about boosting productivity or attracting top talent. It’s about where he thinks the wider economy is moving.

Nvidia clearly expects value to move away from cash and into compute. That’s not a huge leap to make, to be honest. Companies are currently spending billions on AI infrastructure, and demand for GPUs is skyrocketing. As a result, it seems like AI tokens are a pretty good unit of measurement for productivity, which would be incredibly useful for investors.

If tokens become a new standard for measuring work, they’ll get turned into both a budgeting tool and a price mechanism. That could make pricing a lot more accurate, and also effectively turn compute into a commodity. Crucially, Nvidia would be sitting courtside throughout all of this and stand to make a killing.

So, that’s where we’re currently sitting. Nvidia’s wild new compensation idea is still in its infancy, and we’ll just have to wait and see if it actually catches on. But either way, it looks like the writing is already on the wall for the way big corporations define work and productivity. Things are going to start moving very quickly from here on out.