There are few players, but potentially plenty of losers, in so

Anthropic CEO Dario Amodei speaks on Feb. 19 during the AI Impact Summit in New Delhi. Recently, he wrote online that he worries a swarm of artificial intelligence agents “could be capable of taking over the entire internet.”

Anthropic is counting on industry in the U.S. and worldwide to maintain its pace of investing in artificial intelligence.

Nvidia stands to win at artificial intelligence no matter which companies and/or countries come out on top of the so-called “AI race.”

Public angst that artificial intelligence is becoming too big a threat to humans has intensified in recent weeks to the point of dominating cable news shows and podcasts.

The level of attention has accelerated greatly by two occurrences:

Article continues below this ad

Even more recently, lawmakers led by U.S. Sen. Bernie Sanders, I-Vt., and U.S. Rep. Greg Casar, D-Austin, announced they are close to filing federal legislation to ban the building of “artificial superintelligence.”

It’s all transpired amid differing senses of urgency — either an urgency to develop AI as fast as possible or an urgency to contain it just as rapidly.

Like most people, I’m racing to catch up in understanding what’s happened and where this might go. Somewhat ironically, I’m leaning on ChapGPT for answers to the questions that form in my head faster than I can type them into the prompt.

So, while we work through these quandaries, including discussions about creating policy and/or regulations for AI, I suggest that it would be extremely helpful to drop the vague — near meaningless, really — references to the so-called “AI race” we supposedly need to win or else.

While there’s much to unpack in that short imperative, let's focus on two things:

Those in the full-speed-ahead crowd argue our country is engaged in a loser-vanishes-from-civilization battle with China primarily, but also possible rogue nations and actors, for “control” of AI.

Certainly, our military and others are increasingly incorporating AI into daily operations, including weapons systems. Yet it’s odd that multiple uniformed and civilian proponents of modernizing the U.S. military through increased adoption of drones, devices that are undoubtedly made more effective with AI, have been forced out or otherwise diminished.

The more tangible competition is not in warfare but in commerce, where the competitors are corporations and investors in them.

Nvidia is the undisputed global leader in supplying the high-performance, general-purpose accelerators most widely used to develop and deploy advanced AI systems, particularly those used in training AI. While Nvidia does not publicly identify its customers or rank them in terms of sales volume, it’s generally understood that its most lucrative buyers include major cloud and technology providers, such as Microsoft Azure, Meta Platforms, Amazon Web Services, Google Cloud, Oracle Cloud and CoreWeave.

It’s through providers such as these that major AI developers, including OpenAI, Anthropic and xAI, are huge users of Nvidia-powered computing.

So, the more intense the competition among all those companies is, the more Nvidia benefits — and it’s been benefiting a lot. Its second-quarter revenue of $96.2 billion and profit of $59.7 billion were up year over year by 106% and 126%, respectively. And $89 billion of that revenue was related to data centers, up 117% year over year.

Meanwhile, OpenAI and Anthropic — the front-runners, for now, in developing AI models — are not profitable. Although both are privately held, for now, and therefore don’t report earnings like publicly traded companies, financial experts have a pretty good idea of how much money they’re earning and spending. And they need corporate America, including the military, and the rest of the world to keep buying until their balance sheets turn positive.

And private industry is counting on AI to boost productivity, and perhaps reduce labor. The Federal Reserve Bank of Atlanta estimates that AI investment in 2026 among nonfarm industries will be $280 billion, while analysts at Gartner forecast $2.7 trillion in worldwide spending on AI this year and $3.6 trillion in 2027.

The stakes are high, and few of us have chips on the table, at which Nvidia is the dealer.

That brings to the consequences of losing.

To be sure, if U.S. companies dominate AI, assuming the technology doesn’t kill us, our nation stands to enjoy a degree of global economic influence.

But the more likely outcome is that, as with many other game-changing developments, there will be many winners and losers, and they will change over time. Think automobiles, airplanes and air transportation, television, computers, tractors, vitamin supplements. Companies rise, fall and perhaps rise again.

Countries may or may not come along for the ride.

Remember the “space race”? The Soviet Union led for a while. Then the U.S. got to the moon first. Eventually, nations decided to collaborate on exploration, while private industry stepped in to reap the commercial rewards made possible by decades of multinational government investment.

So who won? Many would say Elon Musk.

Still, one big difference with AI, at least according to industry leaders and plenty of lawmakers, is that if we — and here, I mean all of us — do not demand appropriate caution, there will be no winners.