AI Funny Money

Written By Vaclav Vincalek

The first month of 2026 is behind us and the talk about more money for OpenAI is as topical as ever. Tens of billions of dollars are thrown around like chips in the casino. But the nagging question about revenue, and dare I say about profit, is coming up more and more.

The big tech companies are getting scrutinized for their AI investments, investments going to the AI startups to fund the data centers, AI chips and AI compute. Search for the term 'circular AI economy' and you might come across this (and many other) articles “What does the disappearance of a $100bn deal mean for the AI economy?” 

Slowly and surely there will have to be a road to profitability. Asking about revenue in the past was always dismissed with scoffs that one doesn't understand the new amazing, life altering technology and that in the past similar investments had to happen to leap forward. Examples of building railroads, over building fiber cables or the building of Amazon Cloud (AWS) are cited frequently. One can debate the merit of these parallels, but the utility of these investments were obvious - move things from A to B, abundance of bandwidth for fast communication or a global compute scalability.

But what about AI?

Here we are presented with the 'imagine the possibilities' scenario, where the ordinary brain is not able to comprehend what's coming. Based on all the predictions coming from the CEO of OpenAI, we know that “The Singularity is Here, and it's Disappointingly Boring” and “AI Has Already Gone Past The Event Horizon But No Worries Since AGI And ASI Will Be A Gentle Singularity” or “ChatGPT is PhD level smart.”

All that is nice, but still it doesn't answer where the money will come from.

Let's review what OpenAI has tried so far:


The estimate coming from OpenAI is that by the end of 2025 it will reach $20 billion in revenue. Hard to say if API revenue is profitable, but the rest of the revenue is money losing revenue. And sadly, the expenses are climbing faster than the revenue.

Maybe the thought of the senior management at OpenAI is that they will make up the difference in volume.

Does OpenAI have other ideas?

Of course. The most recent one is the introduction of advertising inside the free or low cost ChatGPT coming early in 2026. How does that advertising align with the idea of AI open to all of humanity? You would be happy to know that OpenAI quickly aligned their mission statement - “Mission alignment: Our mission is to ensure AGI benefits all of humanity; our pursuit of advertising is always in support of that mission and making AI more accessible.” And “Answer independence: Ads do not influence the answers ChatGPT gives you. Answers are optimized based on what's most helpful to you. Ads are always separate and clearly labeled.”

What can one expect from this?

If we judge by revenue generated by Google ads, which brought almost $350 billion in 2025 it sounds like a no-brainer. Except, Google with Gemini is already working on it and Google has all the infrastructure and people in place. And OpenAI is losing the audience to Google while Mr. Altman realizes that Gemini is getting better than ChatGPT.

What a pickle. What to do? What to do?

Here is an #idea. What about if all the PhD level geniusness can be used to make money? Since the OpenAI team is running out of ideas, why not use the people who are using the service for free (the ungrateful freeloaders) and crowdsource their ideas? And because we are supposed to operate in the realm of 'imagine the possibilities', imagine this scenario.

You get an idea and want to validate it by using your pal ChatGPT. You start with 'I have an idea for this innovative product, which would do X. Tell me how to validate the idea and turn it into a successful company.' Say no more, ChatGPT goes to work. It starts querying all its knowledge, runs multiple queries through various databases and it 'realizes' that this is something which is not known and can be even patentable. It spins another process which takes the idea further and before you know it, it creates a patent application and sends it to the patent offices around the world. All that while it tells you that it is thinking about it.

Then it provides an answer and it either tells you that the idea is not that great or gives you the option to become a co-inventor on the patent. Considering that ChatGPT has somewhere between 800 million to 850 million weekly users and only 0.001% of ideas are innovative, we are still talking about 8,000 patents a week. What would be required? A simple change in Terms of Use where the output will be 'shared' between you and OpenAI. Impossible? Well, we already did the 'mission alignment' to use the ads to better serve humanity, didn't we? Before you know it, OpenAI will become a patent troll.

The OpenAI team doesn't know if they want to be an infrastructure company and provide services like AWS, or if they will develop products either for consumer or business markets. As they are trying to decide that, they have to compete with everyone else: Google, Anthropic, Meta, xAI or China. The revenue generated by the API access is a blessing and a curse at the same time. It makes money, but it is easy to switch from one provider to another one. While there are other money-making opportunities, none of them generate enough revenue and quickly get them to profitability which will demonstrate a path forward.

Recurrent pattern? Big companies are using OpenAI to pass money back to themselves and keep the illusion of billions in investment going. OpenAI has no other choice than to keep the illusion going and keep using all its resources on money losing ideas. Once the last investment dollar is paid back to AWS, Microsoft or Nvidia for the use of Nvidia chips in AWS and MS data centers; and OpenAI's market share dips below 30% or 20% (it was 70% in Jan 2025 and 45% in Jan 2026), it will become minor market player ready to be bought for cents on a dollar. Any attempt to go public will re-enact the WeWork fiasco.

Vaclav Vincalek
Next

Amazon and the lure of retail