AI isn’t a Bubble

Not personalized financial advice - your money, your choice

Not generated by AI

20 Billion. 

That’s how much revenue OpenAI reported for fiscal year 2025 at the beginning of 2026. Sam Altman says that in 2026, they may surpass 100 billion.

A couple of months back, I wrote an article about circular financing and how AI infrastructure and AI software companies were effectively funding each other.

At the time, this wasn’t too dissimilar to what happened in the early 2000s with internet infrastructure and internet service companies.

Since that article, the collective market caps of AI-related companies have surged some 650 billion dollars. This includes companies: Alphabet, NVIDIA, AMD, Microsoft, Broadcom, Super Micro, SNDK, BNAI, MTC, xAI, OpenAI, Anthropic, Databricks, and Mistral AI.

So… maybe we aren’t in a bubble? The main distinction here between internet and AI companies is that most of these companies (specifically on the software side) are making a ton of revenue. 

In the dot-com bubble, roughly 28% of companies that went public were pre-revenue. They just wanted eyeballs.

The earnings gap was even wider. A whopping 76% of companies that went public in 1999 were reporting negative earnings. If you had a company with ‘.com’ in the name, you could make big money.

Today, the number of pre-revenue publicly traded AI companies is basically 0.

OpenAI, Anthropic, and Perplexity are raising money via venture capital funding and staying private as long as the VC money lasts.

Most of the AI-related companies that the public can invest in right now are large, high-capital-expenditure companies like Microsoft, Oracle, Nvidia, Google, and Palantir. All of these companies aren’t only post-revenue, they’re some of the best profit machines of all time.

The argument here is that the only reason it feels like a bubble is that valuations are sky-high. 

When you buy Nvidia, you’re buying a share of the ‘shovel’ company, and when you buy Microsoft (a significant portion of OpenAI), you’re getting the ‘mining’ company.

If there is a real bubble, it’s obscured by the private credit and venture capital markets, where the private AI companies have been raising money.

Earlier in February, there was a correction that impacted many tech stocks. 

The market thinks that AI companies will crush traditional software companies. As a result, many AI plays remained less affected by the sudden decline.

As AI gets better at editing and creating photos and videos, companies like Adobe will suffer.

Adobe offers a suite of tools for creatives. Products like Illustrator (marketing and branding), Photoshop (photo editing), and Premiere Pro (video editing) make up the tech stack used by marketers, editors, photographers, and videographers.

While the majority of creatives use Adobe products, it doesn’t mean they will continue to. 

My prediction: as AI makes it easier for more people to do those jobs, they will become more competitive, and there will be fewer seats as people in other roles take on that work. 

While this would reduce Adobe’s hold on the creative market, Adobe has been investing heavily in creating AI features. 

Whether they are left behind remains to be seen.

Weirdly, Microsoft was also impacted by these declines despite its ownership stake in OpenAI. 

The market may have forgotten this interesting fact, or it’s worried that OpenAI will start to cannibalize its application business. 

Prediction: Microsoft would sooner add LLM integrations to those products than lose them altogether... but investors disagree.

If this helped you understand why we may not be in an AI bubble, please consider subscribing to my YouTube channel or newsletter! 

Subscribe!