artificial intelligence is entering its fastest growth phase yet. Wall Street isn’t arguing with that. Instead, it’s asking a different question: Who will keep paying for it?

In a post on X over the weekend, the Tesla, Inc. ) and xAI CEO described AI as a “supersonic tsunami,” sharing a chart that appeared to show exponential growth in AI adoption after a brief slowdown in late 2024.

AI is a supersonic tsunami

The post reinforces a view shared by many AI bulls — that the industry’s expansion is still in its early innings. But as Big Tech prepares to spend hundreds of billions of dollars on AI infrastructure this year, investors are becoming increasingly focused on whether that spending can continue at its current pace.

Wall Street’s Debate Has Shifted

The AI conversation has evolved dramatically over the past two years.

Initially, investors debated whether generative AI would become a transformational technology. That quickly shifted to identifying the biggest winners, sending companies like Nvidia Corp. ), Broadcom Inc. (NASDAQ:AVGO) and Palantir Technologies Inc. (NASDAQ:PLTR) soaring.

Now, the focus is changing again.

Rather than questioning AI’s long-term potential, investors are increasingly scrutinizing the economics behind the buildout — particularly the billions of dollars hyperscalers are committing to new data centers, chips and networking infrastructure.

Read Also: Meta's AI Is 'Justifying Higher Capex'—Yet Amazon Gets The Benefit Of The Doubt

The Capex Question

David Russell, Global Head of Market Strategy at TradeStation, believes the next phase of the AI trade may hinge less on AI demand itself and more on investor tolerance for infrastructure spending.

“Nvidia now faces questions about the longer-term demand for AI infrastructure because it’s a bellwether for the entire trend,” Russell told Benzinga in exclusive comments shared over email. “While there are reasons to be bullish on the AI buildout over the coming years, weakness in complementary products like data storage may reflect worries that demand will slow in the coming quarters.”

Russell added that there “could be a pause before AI takes over the world.”

According to Russell, the single most important metric for AI investors isn’t Nvidia Corp‘s (NASDAQ:NVDA) quarterly revenue or GPU shipments — it’s hyperscaler capital expenditure.

“Hyperscaler capex is the main indicator for Nvidia’s business,” he said. “However, investors are increasingly punishing companies for spending on datacenters. That could make the market price in capex cuts before they’re announced.”

AI Isn’t the Debate Anymore

That tension captures the current state of the AI trade.

Musk sees exponential adoption and an industry that is only beginning to accelerate. Russell doesn’t necessarily disagree. Instead, he argues that markets may be growing less patient with the enormous price tag required to sustain that growth.

For investors, the debate is no longer whether AI will reshape the global economy. It’s whether companies building the infrastructure—from hyperscalers to chipmakers — can generate returns quickly enough to justify the unprecedented wave of spending.

As the next round of earnings unfolds, that question may prove just as important for AI stocks as any revenue or profit beat.

Read Also: Alphabet, Microsoft, Meta Earnings Will Spotlight AI Spending — But the Bigger Story May Be In The Footnotes

Photo: Frederic Legrand – COMEO/Shutterstock