The Big Problem For AI Companies — They Don’t Have A Monopoly
I ended my last article with these lines: “Listen to cousin Ed and the gang talk about this for a few minutes in the video below. I’m going to move on to writing the second part of this two-part series, primarily because I think it’s a really big second point that is getting very little attention. Frankly, I think it’s the most important part, and it’s sort of funny that it’s getting ignored so much by the Big Money community since it’s a serious basic of their whole world.”
I’m going to cut to the chase first. The title says it. The big, gigantic problem Big Tech companies have is that none of them have a monopoly on AI.
I’ve been in a variety of discussions and sessions this week that either centered on AI or touched on it. I’ve also been using it a bit for a few things to see how much it can help. One thing that struck me is that while it might have seemed for a while like one or two companies were on a revolutionary path and had a monopoly on the market, the issue is that there are now several companies doing essentially the same thing and there’s really no hope of developing a monopoly. Maybe it’s also because it’s summertime and my kids have been playing Monopoly a lot, but the fact is that this is a fundamental way a company makes a fortune. Google became gigantic because it basically monopolized the “search” market. Microsoft — well, we know all about that. (Those of us who grew up before the turn of the century at least.) Facebook came to dominate social networks in the early stages of that industry. YouTube found a way to edge out the competition and take a completely dominant position in the online video entertainment space. Yes, there are others in each of their fields and there were competitors early on that couldn’t compete, but there were user nuances and network effects that made the winners dominant.
With AI chatbots (LLM companies), there are now several that do the same thing. You put in your text prompt and get your response or document or whatever. You can use OpenAI’s ChatGPT. You can use Google’s Gemini. You can use Anthropic’s Claude. You can use Microsoft Pilot. You can use SpaceX’s Grok. I’m sure there are others I’m missing here, but you get the point. None of them have a monopoly or even importantly dominant position in this field, because, for most uses, how much different are they really? Also, could we not do things before that they can do? They make processes quicker and more efficient (as long as they don’t totally screw things up), but they are doing what people already did, so why can’t option Z do what option A does?
The problem with not having a monopoly or a seriously dominant position in a market like this is that it’s then hard to squeeze out a profit, let alone a massively wealth-creating one.
These companies are all burning billions of dollars of cash, and they are supposedly going to make that back with some combination of subscriptions and ads. However, if one of them tries to charge more (or simply charge anything), users can just say, “Heck no I’m not paying for that — I’ll just use this other option.” So, every time one of them tries to make money to make a profit, they get undercut and lose users. How do they ever make a profit?
Naturally, if several shut down or go bankrupt, eventually, the one that lasts the longest could come out on top. Maybe. But then how much money will have been incinerated by then? How much money will have been invested to outlive the others? And what about those much more efficient systems that get developed as time goes on? What if AI like this just gets commoditized?
As others have said, the dot com boom and bust didn’t mean the Internet collapsed. It’s just that a ton of money was invested into things that weren’t ever going to provide a good ROI, or any ROI. How much money is currently being invested into OpenAI, Anthropic, Google, Microsoft, SpaceX, and others with the dream of “winning” the AI race? How much of that money is ever going to get a positive return from those investments?
Incidentally, just after writing this article, I ran across this video:
Interesting points….
Is there a part 3 to this series? Should I go somewhere else with this topic right now? Is there some great content on this that I’m missing?
UPDATE: I don’t normally watch and see videos in this field, but covering it today has fed more interesting videos to me that touch on some of the things I wrote about above and in the previous article, as well as other related matters: