Analysts see AI growth potential but warn private market exuberance could test investor discipline
AI valuations have surged again, reigniting debate over whether the market is entering bubble territory.
A recent Goldman Sachs ‘Top of Mind’ report argues the sector isn’t yet in a speculative frenzy although the fault lines between public and private markets are widening.
The Wall Street firm’s Eric Sheridan and Kash Rangan contend that while parallels to the dot-com era exist, fundamentals remain stronger today. Sheridan points to the “circularity” of AI investment -model developers, infrastructure providers, and hyperscalers all buying from and investing in one another - as a reason for caution. But he notes that today’s market leaders are financially sound: “most of the Magnificent 7... generate outsized levels of free cash flow and engage in stock buybacks and pay dividends, which very few firms did in 1999.”
Rangan adds that software valuations remain restrained and many companies are funding expansion from cash flows rather than speculation. His bigger worry lies in the rise of debt-driven AI spending, as firms like Oracle and CoreWeave issue major bond sales to finance infrastructure.
Peter Oppenheimer, Goldman’s chief global equity strategist, echoes that the market shows “similarities but key differences” from past bubbles, citing strong balance sheets and genuine profit generation. Ryan Hammond, a senior US strategist, agrees the sector is not yet frothy, outside niche corners such as quantum computing.
However, some cited in the report see storm clouds forming, such as Sequoia Capital’s David Cahn who warns that the scale of data-center construction (possibly reaching $4 trillion by 2030) is hard to justify without the advent of artificial general intelligence. “If you believe there's a data center bubble and there's going to be an overbuild of capacity, then you want to invest in consumers of compute,” he said, arguing that overcapacity will benefit software firms that use AI rather than build it.
By contrast, Bessemer Venture Partners’ Byron Deeter dismisses the idea of excessive spending. He believes current valuations reflect the real opportunity ahead. “This isn’t a ‘hope-and-hype’ cycle like the Dot-Com Era,” Deeter said. “AI is reshaping businesses in ways previously unimaginable.”
Goldman’s economist Joseph Briggs goes further, estimating that generative AI could generate $20 trillion in global economic value, including $8 trillion flowing to US companies. He sees AI investment as “sustainable” given its long-term productivity potential.
Not everyone is convinced. NYU’s Gary Marcus remains a prominent skeptic, arguing that “generative AI is still essentially autocomplete on steroids” and far from true intelligence. He warns the industry has overpromised on productivity gains that have yet to appear.
AI still offers compelling long-term opportunities, but concentration risk is growing. Goldman’s Oppenheimer advises maintaining balance across sectors and regions. He notes that tech remains attractive, but diversification “has paid off this year,” and will likely continue to do so if hype outpaces reality.
