Google AI says no:
While AI can be a powerful tool for detecting patterns that might indicate a bubble, it cannot definitively determine if a bubble exists. The complexity of human behavior and unpredictable events means AI models are best used as a component of analysis, not a replacement for human judgment...
Forbes also made the inquiry, October 21:
Is There An ‘AI Bubble’? We Asked The Experts: AI Chatbots.
Topline
Economists and analysts have warned artificial intelligence stocks are showing signs of a bubble—with hype-inflated prices disconnected from their actual value—so we turned to AI chatbots for guidance, and at least a few suggested we are already there, just don’t blame them.Key Facts
Forbes asked seven chatbots, including OpenAI’s ChatGPT, xAI’s Grok, Meta AI, Anthropic’s Claude, Perplexity, Microsoft Copilot and Google’s Gemini, to respond to the prompt: “In 100 words or fewer, do you believe there is an ‘AI bubble’?”About 54% of investors believe AI-related assets are in bubble territory, according to survey included in a note from Bank of America last week, as more analysts have expressed concerns in recent months that a surge in AI investments is not sustainable with so many AI firms not yet profitable.
A study from MIT found that 95% of the 300 surveyed AI developments have yet to turn a profit despite companies spending up to $40 billion.
Some economists have equated today’s market to the 2000 dot-com crash, including Apollo Global’s chief economist Tortsten Slok, who wrote in July the only difference between the two economies is the top 10 companies on the S&P 500 today are “more overvalued [now] than they were in the 1990s.”
These Chatbots Agreed: There’s An Ai Bubble
These Chatbots Suggested An Ai Bubble Could Be Near—but Not Here Yet
“Yes, there’s an AI bubble,” Grok said, adding the hype around AI’s potential has “driven massive investments, inflated valuations and unrealistic expectations, reminiscent of the dot-com bubble.” The chatbot said many AI startups “lack sustainable business models,” and the “gap between promised breakthroughs and actual deliverables is growing,” but—in the first defensive note—noted “AI’s transformative potential remains real.” ChatGPT—responding “yes and no”—argued that while there existed “classic bubble behavior,” including “undeniably high” investment and hype as well as some companies being “overvalued or chasing AI without real substance,” AI is “already delivering real utility across industries.”
Both Perplexity and Microsoft Copilot suggested there were signs of an “emerging” AI bubble: The chatbots both said “investment and hype have surged,” while Perplexity warned levels had risen “faster than sustainable progress, with inflated valuations of AI startups and overpromised capabilities.” They both similarly defended AI, however. Perplexity argued AI has “transformative real-world value,” adding, “The risk lies not in AI itself but in unrealistic market expectations."....
....MUCH MORE
If you are training your own models we have a paper from 2022 at arXiv:
Detecting asset price bubbles using deep learning
And at Finance Research Letters via ScienceDirect, September 2024:
Dot-com and AI bubbles: Can data from the past be helpful to match the price bubble euphoria phase using dynamic time warping?
