SouthernWorldwide.com – A recent sell-off in technology stocks has highlighted a growing concern among investors: the potential for a significant investment in artificial intelligence to not yield the expected returns if demand falters.
The Nasdaq Composite Index has seen a decline of nearly 3% this week. This downturn reflects Wall Street’s apprehension about whether the vast sums being invested in AI will translate into the revenue and profit growth necessary to justify the enormous costs.
Goldman Sachs projects that technology companies will invest a staggering $7.6 trillion by 2031 to construct numerous data centers required to power the advancement of AI. However, emerging data is casting doubt on whether a sufficient number of consumers and businesses will be willing to pay for these services.
This is particularly concerning as the major tech firms leading the AI charge are increasingly relying on debt to finance the development of this essential infrastructure. Kate Brennan, associate director at the AI Now Institute, expressed this concern to CBS News.
“There’s concern around how much hyperscalers are turning to debt markets in order to finance the infrastructure buildout,” Brennan stated, referring to companies like Alphabet, Amazon, Meta, Microsoft, and Oracle. These are the giants driving the massive capital expenditure in AI.
Brennan further noted that the anticipated returns are not yet materializing, and the claimed improvements in efficiency or productivity are not proving to be substantial enough. She also pointed to a growing skepticism among consumers and workers regarding the practical utility of AI.
While Americans are indeed using AI more frequently, a relatively small number currently appear willing to pay for it. This reluctance is compounded by public apprehension about the technology. According to Pew Research, 40% of adults perceive AI as a negative societal force for the next two decades, compared to only 16% who believe it will be positive.
Simultaneously, an increasing number of companies are opting for AI solutions over human labor, leading to layoffs and fueling concerns about AI’s impact on employment. For businesses, the return on investment from these AI implementations remains uncertain.
A May study by Gartner, a tech research firm, indicated that companies replacing workers with AI agents often fail to achieve a return on their investment. This suggests that the perceived benefits may not always translate into tangible financial gains for businesses.
Brennan suggests that many consumers are interacting with AI not out of a desire to engage with chatbots, but rather because the technology is becoming ubiquitous. For instance, a simple Google search now often yields an AI-generated response at the top of the page.
Similarly, when calling a company’s customer service line, it’s increasingly common to encounter an AI agent, sometimes accompanied by simulated background noise to mimic human activity. This pervasive integration is driven by factors other than direct consumer demand.
“The current push for AI adoption that we’re seeing is directly coming from the financial incentives of AI firms,” Brennan elaborated. Given the immense capital expenditures involved, major AI companies are actively promoting AI integration across various sectors, irrespective of existing demand or customer preference.
Bubble or bust?
Wall Street has long harbored fears of an AI bubble, especially as companies like Alphabet and chip manufacturer Nvidia have repeatedly driven the U.S. stock market to record highs. Some investors are drawing parallels between the current situation and the dot-com bubble of the late 1990s.
During the dot-com era, many early internet companies failed, but those that persevered, such as Amazon and Google, eventually evolved into profitable businesses and widely recognized brands. The AI landscape is expected to follow a similar pattern of mixed outcomes.
Qian Wang, global head of capital market research at Vanguard, and senior global economist Kevin Khang, commented on this in a recent report. They stated that some companies are likely to emerge as highly profitable with significant competitive advantages, while others may find their core businesses rendered obsolete in the evolving AI economy.
As the practical economics of AI continue to unfold, including the trajectory of capital expenditure, the effectiveness of hyperscalers in monetizing AI investments, and the size of the AI market, market volatility is expected to be considerable. Investors should brace themselves for a potentially turbulent period.
The payback test
A critical question regarding the high valuations of hyperscalers and other AI companies is whether their capital spending plans are based on realistic revenue projections, according to economist Ed Yardeni of Yardeni Research.
Companies such as Alphabet, Amazon, Meta, and Microsoft are making substantial investments in data centers and chips, anticipating strong demand for AI services. Concurrently, developers of large language models like OpenAI and Anthropic are incurring costs to utilize these data centers.
However, it remains uncertain whether consumers and businesses will ultimately generate sufficient revenue to justify these significant investments. “The AI ecosystem falls apart if the expected end-user demand for the AI/LLM products does not materialize or if prices for their offerings fall sharply below expectations,” Yardeni cautioned in a note to investors.
Yardeni’s team conducted an analysis of annualized revenue estimates for OpenAI and Anthropic. Their objective was to determine if these companies are acquiring users at a pace sufficient to cover their spending commitments with hyperscalers. This is what he refers to as a “capex payback test,” designed to assess the industry’s ability to sustain capital expenditures.
Their findings indicate that, at present, the AI ecosystem is not fully supported by end-user revenue. However, the situation is not entirely speculative. Yardeni noted that projected revenues for 2030 present a more favorable financial outlook.
These forecasts, however, rely heavily on a significant assumption: AI revenues must continue to scale, and/or compute efficiency must improve. Without these developments, the current investment model may face challenges. “We find that the AI ecosystem is not fully end-user revenue-backed yet, but it is not entirely speculative either,” Yardeni concluded.
