Is the AI Bubble Real? A Wall Street Veteran’s Cold Take

Opening: I’m Not Being a Killjoy, I’m Just Doing the Math

Lately, my social feeds have been hotter than a summer BBQ joint, with someone posting every other day about “this AI company just raised billions” or “that model just broke another barrier.” As someone who’s been around Wall Street for over a decade, watching all this euphoria scroll by, I can’t help but speak my mind:

Folks, we need to pump the brakes.

I’m not here to trash AI or say it has no future. Just like in 2000, the internet really did change the world—but that didn’t stop a massive internet bubble from leaving a lot of people holding the bag.

Today, let’s talk about whether this AI craze has reached the point where we should be hitting the brakes.

AI Is Great, But How Great Exactly?

Let me be clear—I’m not one of those people who dismisses AI outright. Honestly, ChatGPT has made my work life easier in many ways. Self-driving cars aren’t perfect yet, but they’re definitely improving. Translation, document processing—AI handles these tasks better than most interns.

Can AI automate white-collar work? I believe it. Can it accelerate scientific research? Sure. But those folks constantly claiming “AGI is just around the corner” or “humans are about to be surpassed”—I think they’ve been watching too much sci-fi.

It’s like your neighbor Bob who makes decent dumplings. You wouldn’t say he’s going to open a nationwide chain tomorrow, right? Technology develops according to patterns, not according to how big our imaginations can get.

The Power Problem: AI’s Achilles’ Heel

Here’s something most people haven’t noticed: electricity.

You think AI is just writing code and chatting? Wrong. Training and running AI is a real “power hog.” A large AI data center consumes as much electricity as a small city.

Here’s the problem: America’s power grid wasn’t exactly abundant to begin with, and now it suddenly needs to support all these AI data centers. It’s like a car designed for 4 people suddenly needing to fit 8—something’s got to give.

I saw a news story recently where a cloud computing company wanted to build a new data center, and the power company basically said: “Sorry buddy, we don’t have that much juice for you.” Awkward much? The most advanced AI is just expensive scrap metal without electricity.

This is like the real estate boom when land supply couldn’t keep up. What happened in the end? The bubble popped.

Following the Money: How Is This Different from 2000?

As someone in finance, what I care about most is the money trail.

The 2000 internet bubble was mainly overheated equity investment—banks didn’t lend much, so while the stock market crashed, the financial system didn’t collapse. The 2008 subprime crisis was different—banks had too much leverage and nearly dragged the entire financial system down.

Which one does the current AI boom resemble?

So far, AI companies are mainly funded through equity, and while banks and insurance companies are involved, they haven’t gone “all in” like in 2008. That’s good news.

But the question is: if the AI bubble really bursts, can those banks and insurance companies that invested heavily handle the hit? That’s still unclear.

Where’s the Regulation? What’s the Government Doing?

This brings us to an even more practical issue: regulation.

In theory, the government should intervene early to prevent bubbles from getting too big. But what’s the reality? Current governments probably haven’t fully figured out what AI even is—you expect them to create effective regulatory policies?

Plus, AI is now the “engine of economic growth.” Which politician dares to stand up and say “we need to cool down AI”? That’s asking for trouble.

Just like in 2006—how many politicians dared say real estate was overheated and needed cooling? Everyone was afraid of offending voters, afraid of being labeled as “obstructing economic development.”

What Should We Do?

After talking about all these problems, what should ordinary people do?

First, stay sober. AI definitely has a future, but don’t let all the hype go to your head. Invest carefully—don’t bet your life savings on a single concept.

Second, watch the infrastructure. Whether power, networks, and data centers can keep up with AI development demands is more practical than wondering “when will AGI arrive.”

Third, prepare for winter. If the AI bubble really bursts, companies with real technical strength and actual use cases will survive, while those just hyping concepts will disappear. That’s the natural law of technological development.

Final Words

I’m not trying to trash AI or stop people from embracing new technology. I just think that as ordinary people, we need to maintain some calm amid the celebration and keep some rationality in our enthusiasm.

Technology really is changing the world, but the process of changing the world is never smooth sailing. In 1999, everyone said the internet would change everything—they were right, but those who lost everything in the 2000 crash probably wouldn’t consider “changing the world” something worth celebrating.

Will AI change the world? Probably. Will the process be smooth? History tells us probably not.

So embrace change, but don’t forget to buckle your seatbelt.

Follow DreamBeast Programming WeChat for more rational tech insights.