Nvidia Is Riding the AI Revolution. What the Chip Maker Isn’t Telling Us.

Nvidia CEO Jensen Huang continues to tout an AI-powered “industrial revolution”. (AFP via Getty Images)

Cumulative spending on artificial-intelligence infrastructure could hit $4 trillion by the end of the decade, Nvidia said as it posted another set of solid quarterly earnings.

That is an enormous number, even if the earning themselves weren’t spectacular by the chip maker’s lofty standards. Just one critical question remains unanswered.

The word’s biggest AI chip maker has become, of course, the world’s most valuable stock over the past three years. It is now expected to generate more than $200 billion in annual revenue with a profit margin just north of 70%.

The staggering sums, however, pale in comparison to what CEO Jensen Huang sees as the level of AI spending over the next five years.

“We are at the beginning of an industrial revolution that will transform every industry,” Huang told analysts on a conference call. “We see $3 trillion to $4 trillion in AI infrastructure spend by the end of the decade.”

A $4 trillion tally would be bigger than the entire gross domestic product of Great Britain, and largely in line with the combined economic output of Canada and Brazil.

But what Huang wasn’t able to tell investors is how the trillions in spending will translate into bottom-line profits for the non-tech world. None of the megacap tech companies have been able to articulate that either.

To be fair, that isn‘t really his problem. As the “pick-and-shovel” supplier to the AI gold rush, Nvidia is going to make a lot of money regardless of how this new “industrial revolution” plays out.

The revolution Huang envisions also doesn’t have to be that broad. Just three of Nvidia’s biggest customers comprise 56% of its revenue. Keeping them happy, and regaining access to China, will ensure Nvidia’s wheels remain firmly on the soaring-profits track.

Looking at the real economy, however, things are less clear. A recent study from MIT Media Lab showed that most organizations deploying AI technologies are “getting zero return” from their effort.

“Just 5% of integrated AI pilots are extracting millions in value, while the vast majority remain stuck with no measurable P&L impact,” the report said.

The eye-watering sums being forecast for AI spending, and the enormous amounts of infrastructure that is expected to create, won’t make AI any more appealing for the average non-tech company, the study said.

“The core barrier to scaling isn’t infrastructure, regulation, or talent. It is learning,” the study suggested. “Most GenAI systems don’t retain feedback, adapt to context, or improve over time.”

A spring study from S&P Global Market Intelligence appeared to support that view. Around 42% of companies dumped their new AI projects last year, more than double the pace of 2023. The study cited “resistance from customers and employees [and] concern surrounding reputational damage.”

Investors outside of the megacap tech cohort aren’t really seeing enormous near-term gains from AI, either.

Looking at the forward price-to-earnings multiples of the S&P 500’s various sectors, none beyond those influenced by big tech stocks suggest they are ready to outperform their recent past. And of those, only industrials are priced more richly than the benchmark itself. Those valuations are likely tied to the building of AI data centers, not the adoption of the technology.

Early projections of labor market disruption, mass layoffs and the eventual redundancy of entry-level jobs, have yet to be borne out. A study from the Economic Innovation Group found the opposite to be true: Workers that are most exposed to AI replacement are doing better in the job market than those who aren’t.

“By the most obvious measure, then, the effect of AI on jobs is invisible,” the report said.

There’s no question that AI is transforming the tech space. The energy and infrastructure it depends on will improve performance in those sectors as well.

But the broader AI vision of a world where chatbots dominate corporate activity, boost productivity, and generate huge profits simply isn’t in focus as yet.

That said, Nvidia’s still worth $4.3 trillion.

Write to Martin Baccardax at martin.baccardax@barrons.com