Economists and analysts have warned that AI stocks are showing signs of a bubble (with prices inflated by advertising and disconnected from their true value), so we turned to AI chatbots for guidance, and at least some suggested we’re already there, but let’s not blame them.
Key data
Forbes asked seven chatbots, including OpenAI’s ChatGPT, xAI’s Grok, Meta AI, Anthropic’s Claude, Perplexity, Microsoft Copilot, and Google’s Gemini, to answer the question: “In 100 words or less, do you think there is an ‘AI bubble’?”
About 54% of investors believe AI-related assets are in bubble territory, according to a survey included in a Bank of America note last week, as more analysts have expressed concern in recent months that a surge in AI investments is not sustainable with so many AI companies still unprofitable.
An MIT study found that 95% of 300 AI developments surveyed have yet to turn a profit despite companies spending up to $40 billion.
Some economists have compared the current market to the dot-com crash of 2000, including Apollo Global chief economist Tortsten Slok, who wrote in July that the only difference between the two economies is that the top 10 companies in the S&P 500 today are “more overvalued (now) than they were in the 1990s.”
These chatbots agreed: there is an AI bubble
“Yes, there is an AI bubble,” Grok said, adding that the hype around its potential has “driven massive investments, inflated valuations and unrealistic expectations, reminiscent of the dot-com bubble.” The chatbot said that many AI startups “lack sustainable business models” and that “the gap between promised advances and actual results is growing,” but, in the first defensive note, it noted that “AI’s transformative potential remains real.” ChatGPT, answering “yes and no,” argued that while there was “classic bubble behavior,” including “undeniably high” investment and hype, as well as some companies “overvalued or pursuing AI without real substance,” AI “is already offering real utility across industries.”
These chatbots suggest an AI bubble could be near, but it’s not there yet
Both Perplexity and Microsoft Copilot suggested signs of an emerging AI bubble: chatbots said investment and hype had skyrocketed, while Perplexity warned that levels had risen faster than sustainable progress, with inflated valuations of AI startups and exaggerated promises of capabilities. However, both defended AI in similar ways. Perplexity argued that AI has transformative value in the real world, adding: “The risk lies not in the AI itself, but in the unrealistic expectations of the market.”
These chatbots questioned whether an AI bubble really exists
The existence of an AI bubble is the subject of debate among economists, according to Meta AI, which took a neutral approach from both sides, citing concerns such as “overinflated valuations and potential lack of profitability.” The chatbot argued that other analysts believed that an AI boom is “driven by genuine innovation and the potential for significant productivity gains,” leading many to believe that an AI bubble is unlikely, while citing reports of recent Bank of America survey results. Gemini stated that while it is “debatable whether a full-fledged ‘AI bubble’ exists,” he confirmed that there is a “speculative glut in parts of the market.” Gemini added that there were signs of a bubble, such as “sky-high valuations” for startups with “unproven profitability” and “intense market hype,” but argued that the “large companies” driving the growing industry have “solid fundamentals and cash flows.” Claude agreed that there are signs of an AI bubble, but determined that “unlike previous bubbles (dotcom, cryptocurrencies), AI is already generating real value.”
What all chatbots agreed on: It’s not the AI’s fault
Most chatbots insisted that while investor enthusiasm might be getting out of control, there was nothing wrong with the products. As ChatGPT put it: “There may be an AI hype bubble, but not a value bubble.” Copilot agreed: “The key question is not whether AI is valuable, but whether the current enthusiasm is sustainable.” And Claude added: “The question is not whether AI works, but whether current valuations match short-term profitability.”
Most downplayed how a bubble could affect the economy
The “likely outcome” of an AI bubble is a correction in AI-related stocks, according to Claude, but the “transformative potential” of the technology would soften the impact, causing a “shock rather than a complete collapse” of the market. Other chatbots also claimed that the effects of an AI bubble bursting would not be widely felt, including Grok, who said that “economic corrections” would focus on “eliminating overvalued projects and strengthening viable ones” (Grok did not specify which projects). Excitement around AI will “stabilize,” ChatGPT said, which the chatbot said would leave behind “mature, deeply integrated AI systems as part of everyday life.”
Key background
Several economists and technology executives have debated for months whether the influx of AI investments could shock the market. The Bank of England said earlier this month that it had raised the risk of a “sharp market correction,” arguing that an AI-driven market crash was coming, while warning of potential disappointment over advances in AI. Bryan Yeo, chief investment officer at Singapore sovereign wealth fund GIC, said earlier this month that AI startups were raising record amounts of funding and said any company with an “AI label” would be grossly overvalued as investor hype around the technology persisted. Goldman Sachs economist Joseph Briggs wrote last week that multibillion-dollar investments in AI infrastructure are sustainable, but noted that the “ultimate AI winners remain less clear” as the technology changes rapidly.
This text was originally published on Forbes US.
Follow information about business and current events in Forbes Mexico
