AI Tech Is Tanking, Welcome To The Next DOT COM Bubble

According to MIT, Forbes, and Fortune, 95% of companies investing in generative AI are seeing no real returns. The tech sector is showing serious cracks—NASDAQ and other indexes took a hit this week as investor confidence falters, with fears of another dot-com-style crash looming.

Honestly, it’s not surprising. AI has been stumbling for years—deleting codebases, citing fake legal cases, bungling flight bookings, and even botching McDonald’s drive-thru orders. Let’s not forget Builder.ai, the so-called AI startup exposed for using 700 humans pretending to be AI chatbots to mislead investors. And it’s likely we’ll see more companies caught doing the same as things begin to unravel.

Nvidia lost nearly $600 million in January after China leapfrogged the West with DeepSeek, a powerful open-source AI platform built on a fraction of the budget. Since then, it seems U.S. and European companies have either been playing catch-up, exaggerating their capabilities, or hyping vaporware. The much-anticipated but underwhelming release of GPT-5 is a prime example.

Ironically, one of AI’s fastest-growing sectors now is anti-AI—tools designed to detect or block AI-generated content, especially in education. Even Google’s ad algorithms are penalizing websites using AI-generated content, while favoring their own. The result? Lower engagement, fewer clicks, and more frustration. So what’s AI for—other than rewriting mediocre blog posts and generating creepy images of people with too many fingers?

This week alone, over $1 trillion was wiped off the market due to rising fears of an AI bubble. GPT-5’s flop may have just been the first visible crack. The overhype and premature integration of unproven AI tech is having real financial consequences—not just for investors facing billions in losses, but for everyday users stuck with broken tools.

Just look at Microsoft Excel—once a rock-solid app, now undermined by AI Copilot features that come with disclaimers about inaccurate results. When your AI can’t even do basic math, maybe you shouldn’t have added it in the first place.

Here’s a hot take: maybe Apple’s slow approach to AI isn’t so bad. For all the criticism it gets for “falling behind,” maybe it just looked at the state of AI and concluded most of it isn’t ready for prime time. And frankly, it might be smarter to hold off than to break trusted apps with half-baked features.

What a time to be alive.