"If you're not afraid of AI bubble popping, you should be." --Lilah Aparton, 8th grade
If you're not afraid of AI bubble popping, you should be. Plummeting stocks, increased unemployment rates, and the potential for a recession. Trillions of dollars wiped out. But with all these crazy dangers in mind, what even is the AI bubble?
A “financial bubble” is created when investors over spend or over invest into a certain area, leading to the asset having a price much higher than its intrinsic value. This creates a problem, where the stock can’t meet investor’s expectations and it leads to the bubble “popping’.’ And when the bubble ‘‘pops’’ its stock drastically decreases in value, as it can’t generate enough wealth to keep up.
So why can’t investors prevent this?
Investors get FOMO. Especially in A.I’s case where investors are afraid of “missing out” on its benefits. Look at Tulip mania, or the Dot-com crash, they both were created over competition with other investors, and the fact that it was based in something innovative or new. And most investors want to take any opportunity to make more money. All this leads to AI’s assets being overvalued. Jack Blangy, an economics teacher at OSA says, “It is common for new technologies to be over-invested in and for there to be a crash. This happened with excitement around the internet in the late 90s, and it also happened with [r]ailroads in the late 1800s. The scale of investment in AI is similar to the scale of investment in the dotcom bubble (~1% of GDP). However, many dotcom companies were not making any money at all.”
So if the AI bubble does pop, it may still cause money to be wasted, but it would only slow AI being developed. It wouldn’t destroy the economy or ruin AI data centers beyond repair. Like any other financial bubble, the economy can still recover. “AI companies are generating revenue, it's just not as much as their investment.” Blangy continues, “Obviously the dot-com crash did not destroy the internet, and the internet is more important to our world today than it was in the 90s.”
One of these big investors is Nvidia, one of the biggest AI companies in the world. As of January 2026, they’ve invested 2 billion into Coreweave, an AI data center, with a promise to invest another 6 billion later. They’ve also invested 100$ billion into OpenAI (who owns Chatgpt) according to Yale insights.
What happens if the bubble doesn’t pop?
AI keeps improving, and investors get their big returns. “If the bubble doesn't burst it is because AI companies are making large returns on their investment, meaning people and companies are adopting AI and paying money for AI. In this potential future AI is a large part of our world and has been quickly adopted. AI companies will be able to continue to invest heavily in AI and will be able to develop better models that are more capable,” says Blangy.
Companies like OpenAI, Nvidia and Microsoft will keep developing faster. But AI does have limits, cracks are now showing, with bigger models not yielding proportional gains. Models may be ten times larger than they were a couple of years ago, but they’re not 10 times smarter by most metrics. This puts all those companies at risk because now that throwing more money or computing power isn’t working as well as before, it leads to more money being wasted.
So, besides the possible downsides these companies could have on the economy, they have detrimental effects on the environment too. According to Goldman Sachs one Chatgpt question uses as much power as 10 Google searches. These costs also are another step toward an economic crash. OpenAI relies almost completely on investors because it can’t make its own profit. The electric bills, the water for cooling servers, this all contributes to massive debt for OpenAI and other major A.I companies. OpenAI is predicted to lose 15 billion this year, and won’t make a profit until 2029. Which leads to the question, will the AI bubble pop by then?
A “financial bubble” is created when investors over spend or over invest into a certain area, leading to the asset having a price much higher than its intrinsic value. This creates a problem, where the stock can’t meet investor’s expectations and it leads to the bubble “popping’.’ And when the bubble ‘‘pops’’ its stock drastically decreases in value, as it can’t generate enough wealth to keep up.
So why can’t investors prevent this?
Investors get FOMO. Especially in A.I’s case where investors are afraid of “missing out” on its benefits. Look at Tulip mania, or the Dot-com crash, they both were created over competition with other investors, and the fact that it was based in something innovative or new. And most investors want to take any opportunity to make more money. All this leads to AI’s assets being overvalued. Jack Blangy, an economics teacher at OSA says, “It is common for new technologies to be over-invested in and for there to be a crash. This happened with excitement around the internet in the late 90s, and it also happened with [r]ailroads in the late 1800s. The scale of investment in AI is similar to the scale of investment in the dotcom bubble (~1% of GDP). However, many dotcom companies were not making any money at all.”
So if the AI bubble does pop, it may still cause money to be wasted, but it would only slow AI being developed. It wouldn’t destroy the economy or ruin AI data centers beyond repair. Like any other financial bubble, the economy can still recover. “AI companies are generating revenue, it's just not as much as their investment.” Blangy continues, “Obviously the dot-com crash did not destroy the internet, and the internet is more important to our world today than it was in the 90s.”
One of these big investors is Nvidia, one of the biggest AI companies in the world. As of January 2026, they’ve invested 2 billion into Coreweave, an AI data center, with a promise to invest another 6 billion later. They’ve also invested 100$ billion into OpenAI (who owns Chatgpt) according to Yale insights.
What happens if the bubble doesn’t pop?
AI keeps improving, and investors get their big returns. “If the bubble doesn't burst it is because AI companies are making large returns on their investment, meaning people and companies are adopting AI and paying money for AI. In this potential future AI is a large part of our world and has been quickly adopted. AI companies will be able to continue to invest heavily in AI and will be able to develop better models that are more capable,” says Blangy.
Companies like OpenAI, Nvidia and Microsoft will keep developing faster. But AI does have limits, cracks are now showing, with bigger models not yielding proportional gains. Models may be ten times larger than they were a couple of years ago, but they’re not 10 times smarter by most metrics. This puts all those companies at risk because now that throwing more money or computing power isn’t working as well as before, it leads to more money being wasted.
So, besides the possible downsides these companies could have on the economy, they have detrimental effects on the environment too. According to Goldman Sachs one Chatgpt question uses as much power as 10 Google searches. These costs also are another step toward an economic crash. OpenAI relies almost completely on investors because it can’t make its own profit. The electric bills, the water for cooling servers, this all contributes to massive debt for OpenAI and other major A.I companies. OpenAI is predicted to lose 15 billion this year, and won’t make a profit until 2029. Which leads to the question, will the AI bubble pop by then?