Kong CEO Sees AI Infrastructure Boom Outlasting Bubble Fears

The Infrastructure Bet Behind AI’s Growth

Despite growing concerns about an AI investment bubble, industry leaders are betting that today’s massive infrastructure spending will pay off long-term, even if some companies falter along the way. Kong CEO Augusto “Aghi” Marietti recently told Business Insider that while a market correction seems inevitable, the underlying infrastructure being built will remain critical to the technology’s future.

“We’re in this new builders era where it’s a very singular moment where we are going to probably deploy more capex and more capital for enabling the AI era, and we need it,” Marietti explained. His comments come amid reports that major tech companies could spend an estimated $320 billion on capital expenditures, primarily for AI-related infrastructure.

Energy Emerges as Critical Constraint

Interestingly, Marietti identifies energy availability rather than capital as the most immediate bottleneck. “We don’t have the energy we need to power all the GPUs in the following year,” he noted, echoing concerns that have prompted some AI companies to develop their own power solutions.

This energy challenge highlights the physical constraints facing AI’s rapid expansion. Unlike purely digital technologies of the past decade, advanced AI systems require enormous computational resources that translate directly into massive electricity demands. Industry observers note that companies are now competing not just for talent and chips, but for reliable, scalable power sources.

Historical Parallels to Railroad Expansion

Marietti’s perspective aligns with other industry leaders who see current spending as foundational rather than excessive. He draws direct parallels to 19th century railroad construction in the United States, where initial overbuilding eventually gave way to essential infrastructure that transformed the economy.

“Some railroads were deployed ahead of time, but then all the railroads got used,” he observed. “I think in AI, we’re just deploying ahead of time, and eventually something will blow up for a little bit, but we would eventually need the infrastructure that we’re deploying anyways.”

This historical comparison resonates across the industry. Like the railroads that initially connected major cities before expanding to serve smaller communities, today’s AI infrastructure may seem concentrated in specific applications before finding broader utility.

Wall Street’s Bubble Concerns

Meanwhile, financial analysts are watching the spending patterns with increasing concern. The scale of investment by Big Tech companies and leading AI startups has generated significant bubble talk, with some economists suggesting that current capex spending is propping up the entire US economy.

OpenAI CEO Sam Altman acknowledged in August that AI could indeed be in a bubble phase, though he remains optimistic about the technology’s long-term potential. The tension between immediate financial sustainability and long-term strategic positioning creates a complex landscape for investors and executives alike.

What makes this moment particularly challenging for market watchers is the simultaneous presence of genuine technological transformation and speculative excess. Sorting which investments represent foundational infrastructure versus temporary hype remains the central puzzle for the industry.

Beyond the Potential Downturn

Despite anticipating what he calls “a down moment,” Marietti believes the infrastructure being built today will endure regardless of market fluctuations. “After that, we’ll still use all the infrastructure that we build,” he predicted. “We still use the railroads that we deployed 150 years ago ahead of time.”

This long-term perspective suggests that even if specific companies or projects fail, the underlying computational capacity and network infrastructure will continue to support AI development. The question becomes not whether the infrastructure will be used, but which organizations will ultimately control and benefit from it.

As CEO of a company that helps manage API connectivity across complex systems, Marietti occupies a unique position to observe how AI infrastructure evolves. His comments reflect a broader industry recognition that we’re building the digital equivalent of transportation networks—and like those historical projects, the full economic impact may take years to materialize.