As AI continues to eat the world, a looming question is whether we’re in a bubble. That question doesn’t mean to suggest that AI doesn’t have value. It’s more a matter of whether or not it’s overvalued. As everyone rushes at it, does it collectively get more investment and excitement than it deserves?
This is a question of saturation. AI has an optimal level of supply that is unknown – a level that aligns with its demand and room for applicability. That’s equilibrium – classic macro-economics – and isn’t a static thing. It’s a moving target for markets to avoid leaving demand unfulfilled, yet not overshoot the mark.
The latter gets us back to the question of a bubble. Is all of the FOMO-driven excitement on the supply side – and correspondingly, further upstream in venture funding appetite – more than the market will bear? As always, no one knows the answer, but lots of smart money says that the answer is yes.
The latest evidence comes from OpenAI board chair Bret Taylor. As an astute and prolific tech exec, he admitted that we’re likely in an AI bubble. In an interview with The Verge, he was asked to respond to Sam Altman’s recent comment that “someone is going to lose a phenomenal amount of money in AI.”
“I think it is both true that AI will transform the economy, and I think it will, like the internet, create huge amounts of economic value in the future,” said Taylor in the interview. “I think we’re also in a bubble, and a lot of people will lose a lot of money. I think both are absolutely true at the same time.”
History and Physics
Taylor’s comments should carry some credibility – not just because of his acumen, but his motivation. As OpenAI board chair and CEO of Agentic AI company, Sierra, he has no incentive to undercut AI. This also aligns with a trend towards tech execs being more explicit about AI’s potential negative impacts.
Taylor went on to make an important historical reference: AI’s landscape today resembles, in some ways, the dot-com bubble in the late nineties and early 2000s. Many companies spectacularly failed when the bubble burst, but it doesn’t mean that all of the excitement and investment touted in 1999 were wrong.
In fact, we often look back on that period and point mockingly to its irrational exuberance. But the truth is that all of those dot-com-bubble-era predictions came true, and many were even understated. All those promises about the web and ecommerce fell flat at first but started to come true about five years later.
As they say, being early to markets is one of the worst business decisions you can make. In the dot-com boom, it was a matter of underdevloped infrastructure to support all of those internet companies. But broadband and cultural demand caught up to the web and eCommerce, which then exploded.
So will the pendulum swing in the same pattern for AI – first towards over-investment, then back towards correction? Pendulums eventually end up in the middle due to gravity, which I suppose is a metaphor for the invisible-hand-driven equilibrium noted earlier. But they first have to swing widely to each side.
Header image credit: Lanju Fotografie on Unsplash
