Wall Street’s AI trade is starting to look increasingly split between believers and sceptics.
In the space of just a few days, investors have been confronted with several mixed signals about where the artificial intelligence boom is heading: Nvidia Corp (NASDAQ:NVDA, XETRA:NVD) ramping up its already extraordinary investment push across the AI ecosystem, famed short-seller Michael Burry warning of bubble-like conditions, and veteran strategist Ed Yardeni lifting his S&P 500 target yet again.
Together, the developments capture the growing tension driving global tech markets — whether AI is still in the early stages of a transformational investment cycle or drifting into the kind of speculative excess that has historically ended badly.
Nvidia evolves from chipmaker into AI kingmaker
The clearest sign of how quickly the AI arms race is escalating came from CNBC reporting that Nvidia has already committed more than US$40 billion in AI-related equity investments in 2026 alone.
Rather than simply backing startups, Nvidia is increasingly using its balance sheet to secure the infrastructure needed to sustain AI’s explosive growth — spanning data centres, optical networking and compute providers that ultimately become major customers for its hardware.
Last week, Nvidia announced a deal with Australian-founded data centre group IREN, formerly Iris Energy, where the chipmaker secured rights to buy up to US$2.1 billion worth of shares as part of a broader AI infrastructure partnership. The agreement reportedly could support deployment of as much as 5 gigawatts of AI data centre capacity.
Nvidia has also expanded its relationship with Corning, backing new US optical fibre manufacturing projects tied to AI data centre growth. CNBC reported Nvidia provided multibillion-dollar prepayments alongside an equity investment of up to US$3.2 billion to support new capacity.
The shift highlights how AI bottlenecks increasingly extend beyond semiconductors themselves, with fibre optics, networking hardware and power infrastructure becoming just as critical as GPUs.
Bubble fears return to centre stage
For critics, Nvidia’s spending spree is also becoming symbolic of a market that may be feeding itself.
Analysts have increasingly questioned whether the AI trade is becoming too circular, with major technology companies simultaneously financing suppliers, customers and infrastructure partners while benefiting from the resulting demand surge.
That concern sits at the heart of Michael Burry’s latest warning.
The investor best known for predicting the 2008 financial crisis has argued markets are behaving increasingly like the late stages of the dot-com era, with AI enthusiasm overwhelming traditional valuation discipline.
The scale of capital now flowing into AI infrastructure is adding to those concerns. Data centre developers, cloud providers and networking groups are collectively committing hundreds of billions of dollars to expansion despite uncertainty over how quickly demand will translate into sustainable returns.
Bulls still see an earnings-driven boom
Not everyone believes comparisons with the dot-com crash are fair.
Yardeni this week raised his 2026 S&P 500 target to 8,250, arguing the rally remains supported by genuine earnings growth, AI-driven productivity gains and resilient corporate spending.
Unlike many internet-era companies that traded largely on ambition, today’s AI leaders are already generating enormous revenues and cash flows. Nvidia itself has become one of the world’s most profitable companies as hyperscalers race to build AI infrastructure.
Still, investor sensitivity to funding announcements and capital intensity is becoming more obvious.
IREN’s shares surged after announcing its Nvidia partnership before retreating as markets focused on the enormous funding requirements needed to scale AI infrastructure.
For investors, that may be the clearest sign of where the AI trade now sits: powered by genuine technological change, but increasingly shadowed by fears expectations may already be running ahead of reality.