I think so, too, but not at all in the way Altman thinks it's going to happen.
We've already figured out that you can make smaller simpler models that get like 80% of the benefit for like just a few percent of the resources. Like I've got an AI running on my old 2011 computer, no graphics card, that is mostly as good as the Bing AI or whatever, just kinda slow.
Point is, there is a good chance a lot of people will be using AI for the things it is good for (mostly I use it for programming questions and similar) but that does NOT mean that everybody in the universe will be paying Altman's company bazillions of dollars to do so.
Most people - and even companies - that get decent use out of AI are likely to be running their own AI at a small fraction of the cost. Or buying it at MUCH lower-cost from non-cutting edge type "utility" companies that are not on the cutting edge but provide a perfectly usable product. And only a few of the largest/richest companies & people who have dollars to burn and really need "the cutting edge" for whatever reason, will be paying giant trillion-dollar corporations like OpenAI to get the absolutely latest and greatest.
So could indeed be a bit like electricity, in the sense that it is ubiquitous, cheap, and useful. But that means 1000s or millions or even billions of individual providers, not necessarily one or a few giant sprawling tech firms led by sociopathic assholes like Altman (Zuckerberg, Musk, Bezos, etc.).
Just because something will indeed be valuable in the future does not mean that value is infinite, that we should pour infinite money into it, or that the value people are assigning it right now is correct (see e.g. the Dot Com Bubble and pretty much every economic bubble in the past - the internet became very valuable indeed but not in the ways and not to the companies, for certain, that people at that time thought it would).
In fact, the value that people and markets are now assigning to AI companies is almost certainly greatly exaggerated or misplaced. In part because the very process that is creating an interesting and useful product now will make it vastly simply, smaller, easier, and cheaper in just a few years. And in another part because the ways profit will be generated from AI in the future are not the same as they are now - and thus the companies on the receiving end of it are not necessarily going to be the same.
In short, get ready for the Great AI Crash of '26 or '27 or '28 whatever.
Some companies and technologies will survive it but Altman's may or may not be one of them.