Companies abandon AI as prices skyrocket

Last Updated on 4 June 2026 by Willem Moore



Open AI CEO Sam Altman with a red line behind him

For the past few years, billionaires and their media lackeys have told us that the AI revolution is inevitable. Opposing this narrative, some are warning that the ‘generative AI’ technology in question lacks the competence or cost-effectiveness for such a transition. And the emerging signs are proving those naysayers correct:

Why now?

Over the past few weeks, anyone who follows the AI industry will have noticed a big increase in stories like this:

Why is it happening now? It’s because AI companies have switched from a subscription-based offering to a pay-per-token model. The ‘tokens’ in question are what AI models use whenever they process a request, with more tokens used depending on the complexity of what’s asked.

As it turns out, many of these companies were burning through an unfathomable quantity of tokens:

As the sky falls in, OpenAI CEO Sam Altman is pretending to not understand what’s happening:

Altman does have to say something to reassure investors, but we’re not sure this is it.

Ed Zitron is one of generative AI’s most vocal critics. Responding to the latest developments, he highlighted a case in which one AI user used 50% of their token credits with just one prompt. This is a problem, because some users like coders had grown accustomed to making hundreds or thousands of prompts a day.

As Zitron highlights, companies are now paying the “actual price” of AI.

Up until now, AI companies used multi-billion pounds cash injections from wealthy investors to subsidise their technology. They hoped AI would prove to be so useful that when they switched to a more realistic pricing point, companies would have no option but to carry on paying.

Yeah, so about that …

Expensive? Yes. Useful? No.

To make things worse, companies already weren’t making money:

As CNN reported in 2025:

The artificial intelligence industry has a big problem: 95% of companies that try AI aren’t making any money from it, according to a report from the Massachusetts Institute of Technology last month.

The same article carried multiple quotes from supposed experts who promised AI would be very profitable as soon as these businesses pulled their fingers out and implemented it correctly. A year later, the technology is no more useful yet considerably more expensive.

Many companies laid off employees hoping they could replace them with AI. Many companies have since realised the error of their ways:

Not yet, not this

While it does seem likely the world will one day be run by smart, automated systems, the generative AI that’s being sold to us isn’t ‘smart’. Functionally, it’s little more than a juiced-up auto-type, but the tech chancers claimed it was a digital god that would perform our menial tasks for us.

The problem now is that the global economy is propped up by the over-inflated AI bubble. And when that bubble pops, it won’t be the billionaire tech bros who suffer.

Featured image via Chip Somodevilla (Getty Images)

By Willem Moore





Source link