Why “Free” Could Sink The AI Bubble

24/7 Wall St

Quick Read

AI companies like Meta, Google, and Microsoft have spent trillions on infrastructure with most of it going to Nvidia. Yet over 99% of users access AI tools for free with no clear path to monetization. Companies building vertical specific AI applications for sectors like law or healthcare may capture revenue where foundational AI builders cannot. Some investors get rich while others struggle because they never learned there are two completely different strategies to building wealth. Don’t make the same mistake, learn about both here.

Watch the Video

Too Much Capital, Too Few Paying Customers

I told Lee that artificial intelligence might be the most overfunded “free product” in modern business. We both use AI tools every day, and like most people, we don’t pay a dime. That’s the problem.

Trillions have been spent on data centers, chips, and training models, yet the vast majority of users aren’t customers. Venture-backed firms like OpenAI and Anthropic depend on free traffic and vague enterprise plans. A recent Janus Henderson report said AI monetization is still in “early progress.” Another warning came from InvestingLive, noting that consumer AI “struggles with monetization” and “infrastructure costs remain high.”

The Bubble Risk Is Real

Lee pointed out that this looks a lot like a bubble, citing “the sheer trillions of dollars” going in without visible return. I agreed. As MarketWatch recently put it, we might be watching “a slow-motion AI bubble deflation.” Hedge funds like Elliott Management even called Nvidia’s valuation “bubble land.” (FT)

Where the Real Winners Might Be

I told Lee that I still think there’s opportunity, but not for the foundational AI builders. Instead for the companies that build on top of them. Businesses that design AI tools for specific verticals like law, healthcare, or logistics. All of these sectors could end up being the ones that actually charge and earn. A McKinsey study agrees, suggesting future AI monetization will depend on targeted, value-based services.

Until then, I said it plainly: if people won’t pay for AI, the sector can’t make money, and that’s not sustainable.

Transcript:

[00:00:04] Douglas: Lee, there's an emerging. Opinion that shorting the entire AI sector, that means utilities, it means the big tech companies. It means the financial firms that are, know, funding data centers. It means the venture capitalists that have put money into things like, you know, OpenAI, Anthropic. The theory is basically this AI is mostly free. Huge majority of the people, more than 99 out of a hundred people

Story Continues

[00:00:38] Lee Jackson: Oh yeah. I use it every day for free.

[00:00:40] Douglas: that there aren't any other business models that exist, that have been successful, incredibly successful years and years and years don't have a large number of paying customers either.

[00:00:55] Douglas: Millions of people paying. You know, sort of some money, maybe it's $10 or a small number of enterprise customers

[00:01:04] Lee Jackson: Right.

[00:01:06] Douglas: millions, if not billions of dollars. Right now, the artificial intelligence rev Revolution falls into the first category without any question. Almost everybody using AI doesn't pay a dime.

[00:01:24] Lee Jackson: No, I never have. And I think so can get upgraded platforms, I think on Claude and other ones that are pay platforms, but I mean, I'm sure it's way above my pay grade to even use 'em, so I don't have to. So yeah, I think it's interesting. And plus we're hearing now, Doug, and it's getting louder. And louder and louder, there may be an AI bubble and you know, I mean the, the sheer trillions of dollars that have been thrown at this, I think, I think the big concern across Wall Street and maybe across the world is that for the kind of capital that's being put into this.

[00:02:04] Lee Jackson: It's kind of the road you're going down is where, how is this gonna be recouped? It's one thing to see how Nvidia will recoup stuff. 'cause they sell chips and it's another thing to see how, you know, other people involved in whatever area of the AI explosion, whether it be the utilities, well they sell electricity, you know.

[00:02:23] Lee Jackson: But yeah, just the sheer chat bot, you know, function themselves, how are they gonna make back the kind of money that's been thrown into this? this?

[00:02:33] Douglas: as a publicly traded company, no one has had to put any money into Nvidia in decades. I mean, as a startup, maybe they did for a few years. Um. So you may not be able to, you

[00:02:46] Lee Jackson: I mean, not from a capital standpoint.

[00:02:49] Douglas: But it's an honest to God business. You know, it's a, it's not hard for us to say, a company, here's how it makes money.

[00:02:57] Douglas: You may say, well, the multiple on that, money's too high. But when you move to OpenAI, Anthropic, and then I'm gonna segregate for a second, parts of Microsoft and Meta (NASDAQ: META) that are close to pure plays in AI. So let's take Xbot, let's take all the other stuff that Microsoft does. Okay? The, the businesses that they've been in for years,

[00:03:22] Lee Jackson: Okay.

[00:03:22] Douglas: pre AI, and let's segregate the money they put into AI to turn into businesses. The return on that second pot of money is potentially never. If, if they invest, Microsoft invests a

[00:03:39] Lee Jackson: Exactly.

[00:03:40] Douglas: dollars into AI, say chips, energy data farms, you know how much money you'd have to to get a return on that value over time. If you, if you just said you have to, you have to take

[00:03:58] Lee Jackson: It is gigantic. It has to be.

[00:04:00] Douglas: the year that you put the in. You've gotta make that money back at some

[00:04:10] Lee Jackson: Well, a, again, at least Microsoft, you know, for, for their, foray into the field, they have other income streams, at least Alphabet and Google for their foray, into that field have other income streams. But, but all of the companies that are basically providers, you know, AI providers from a chat standpoint, they can't all survive with, with no fee forever.

[00:04:35] Douglas: Number one. Number two. The, the entire premise that's being challenged right now is, is that there is a way to get people to pay, uh, OpenAI money to use, you know, it, it's primary product, which is free. I mean, I understand you can get an upgrade for 10 bucks a month, but the people who will make the money, here's my projection on products from OpenAI are companies that can make adaptations to that and sell it based on segments, industry segments. take an AI application, I design it to replace appellate attorneys, and you pay me for that software application. But in that case, I added some flavor. Put some salt on it, some paprika. I, I've designed it so that a company can actually use it for a specific set of purposes where it's replacing humans or it's replacing something else, but being the primary provider an AI, uh, software function, people aren't gonna pay for it. You don't, there's no case in businesses I know in history where it's free. It's free. It's free. It's free. Oh God.

[00:05:57] Douglas: You gotta pay for it. And everybody's, oh, we're happy to. Even the internet wasn't free

[00:06:03] Lee Jackson: Yeah,

[00:06:03] Douglas: think, oh, the internet was

[00:06:05] Lee Jackson: That never happens.

[00:06:05] Douglas: money for for the AOL every month get on the internet, and I still pay my cable provider Spectrum. The internet is not free to me. I can't get there without writing a check. write it to the internet, but I do write it to somebody who

[00:06:23] Douglas: allows

[00:06:24] Douglas: me access.

[00:06:25] Douglas: I don't think.

[00:06:29] Lee Jackson: Yeah. Well, and I write, I write, I do the same. I write a check to Comcast via Infinity or Xfinity every month. And, uh, gladly because it's reasonable and my service is great, it's super fast, but, you know, I have to have that, I don't have to have a chat bot. 'Cause, you know, if, if I'm, I can always just, you know, use search normal search and find it myself eventually, just not as fast.

[00:06:55] Douglas: the, this is my final. Say on this, AI will collapse as a business for one reason. They won't be able to charge for it. Okay. There's, it may be it, it the fact that it could be one of the most important advances in human history. I don't even wanna get into that bit 'cause I'm not smart enough. I am smart enough to know that if you spend trillions of dollars building something and then give it away for free, it's not a good business.

[00:07:26] Lee Jackson: No, that's, it's kind of hard to justify that business model. Yeah, I agree. It, it, it's gonna be interesting to see how this plays out because they're gonna have to make sure that their AI applications can fit into corporate, uh, sort sort of situations and how, how can they help, you know, big corporations, because it won't be to tell me, you know, how many doodle dogs exist, you know, or, or whatever. You know, people are asking Chat and, and I know there's a zillion functions, but again, until there's an inbound pay or a pay wall, like so many people use, I, I agree. I don't see how they, how they're gonna monetize this and make money, especially with intense CapEx that's gone into.

[00:08:08] Douglas: than me say that, uh, there's gonna be an AI collapse. I'm gonna say it 'cause I'm stupid enough to simply say, if people won't pay for something, you can't make money on it.

[00:08:16] Lee Jackson: Yeah, we'll kind of leave it at that and see if it collapses down the road.

View Comments