Most things work this way lol
Heavy unprofitable investment with hope that the thing matures with time and expansion resulting in profits.
“Top Economists”
Implies that people are largely still pretty incompetent with it. The pioneers have a lot to win.
Fancy being a top economist and therefore being privy to knowledge that is beyond the capabilities of us mortals. We thought the AI was a profitable business, and it turns out it’s a giant scam, who would have thought it?
If you thought AI was profitable that’s on you. Normal people have had the information required since the start.
It’s also expected that new tech companies won’t see profits for the first 10 years. That’s not new.
Regardless, fuck AI and its speculators. Onward the Butlerian Jihad.
I think AI could have a good tool for specific purposes and also a very lucrative business if only they had not thrown all that money at it.
The extreme investments forced them to roll it out prematurely over a much larger market that it could ever supply in a meaningful way.
They’re trying to force a camel through the eye of a needle.
They have been able to use it on less powerful deceives lately. I’m waiting for something like an ebook reader to be able to run an LLM. Imagine you forget who a character is and the on device AI parses the whole book and explains what you have forgotten.
The problem with capitalism is that morons can get filthy rich and make poor decisions.
Example 1: US president.
There there. Give some credit to the American people. They’re the ones that elected him. Twice.
I mean really the same problem with Monarchy. Once someone somehow gets on top very hard to remove them and their descendants from power even if the original person or their descendants are complete idiots.
Yeah, look at Zuckerberg, who recently burned billions on his crappy virtual reality project, and is now doing the same with AI.
The stupid thing is there are already profitable VR games that basically do exactly what Zuckerberg was trying to do. So despite the fact that he could just copied somebody else’s homework he managed to screw it up.
The goal isn’t consumer level profits, the goal is to replace workers, this is the first time capitalists are looking farther than quarterly profits. They’re investing in R&D to replace us, not selling a product. We are not part of the plan.
Oh my god thank you. Somehow despite being engulfed in this stuff being an I.T person while having affection toward Ed Zitran and full agreement on almost all arguments living this weird ying yang over the topic.
This is the succinct detail I hadn’t clued into until this moment of what the core motivation is for all of the crazy spending that makes zero sense. It’s literally this. There is no amount of money, lying, cheating or stealing that cannot be justified by the most cynical of capitalist if they believe this is the outcome.
It’s such an obvious thing when you put it like that and I’m even sure I’ve seen it said before but it really hit me for some reason from your comment. They would move mountains if they thought there was even the smallest chance and bet the entire farm over it.
First you have recognize there is no barriers to their cynicism I guess but this is the only root cause conclusion that could be rationalized. That or fomo and temporary wins but for those spending and not on any receiving end, I couldn’t make sense of the rational until you said it so plainly.
They are paying for fantasy about rendering people redundant, the sik fuks.
It’s not even capitalism anymore, assuming they expect their peers in corruption to actuallg pay each other and themselves a fair amount.
Together they’re on the hook for over a trillion dollars in lease payments, so far, that haven’t started yet. They’re counting on turning a profit before they need to start paying that back. That’s why they’re shoving it down our throats, they either make this shit work or they go down in flames.
I’m betting they go down in flames, base on the absolute shit-show “AI” has turned out to be. They keep claiming they have fixed it, or they will fix it, and then it shits the bed again in some public, spectacular way.
I bet they get bailed one last time by the government, then people start dying fast.
For some reason, I find this satisfying and funny. The landscape has been changing. The open-weight models from Chinese labs are getting really good and becoming a serious threat to the closed models from American AI companies.
Interestingly, Google, Meta, and Microsoft — who all have revenue streams outside of AI — recognize the threat are getting more supportive about open-weight models. Anthropic and OpenAI support keeping models closed for “safety” reasons.
I think when they start feeling the flames, they’ll also pull us into it. Like selling LLMs meant for exploits and cyberwarfare to the public.
I mean that’s 4 companies though, so we’re like saying “Microsoft has something like 250 billion in leases”… But they are “worth” over 3 trillion and hold assets worth 2-3 times those leases. Don’t think it’s enough to sink those large companies, it’s just general investing at their scale now I guess.
Personally I think they are all overvalued, but unless all their shit flops quickly, they’ll be around in 10 years still.
Investment shouldn’t count as earnings. You’re literally borrowing the money…
Why is it that almost all big financial situations in my lifetime are Ponzi schemes?
This has Ponzi elements, but it’s closer to a multi-company Enron in its structure. They’re manufacturing fake revenues through circular investments. Nvidia invests in AI companies and increases their value while AI companies use that investment to buy Nvidia cards increasing its value so, so they invest more in AI, so they buy more cards…
Enron did this by making fake companies to generate fake revenue. Right now the tech sector is all working together on this scheme, so none of the companies are fake, but the revenue is.
The scheme requires an every increasing amount of new “investment” or the whole thing collapses. Is it more then a ponzi scheme? Sure, most normal schemes can’t do what these companies can do but at the end of the day the scheme is still very much a ponzi structure.
These (non hardware) companies are not relying on fake revenue, they all post massive losses in AI on the regular and yet like some sort of massive lost cost fallacy the markets still keep investing. Maybe next quarter (insert AI) will start making money, but more likely it will all pop before. And the hardware companies making money hand over fist? They are getting bodied on the market (see the South Korean situation).
None of this is even unexpected or seen as odd anymore, the financial systems, markets, private equity and most funds seem to just treat the situation as normal. I have seen nothing but once in a lifetime financial fuckups every few years since I was in high school, and almost every one is in the end is just some ponzi style scheme that was always doomed to collapse.
The key difference with a Ponzi scheme is that the Ponzi/Madoff investors weren’t in on the schemes. They though they were investors in a legitimate enterprise. They had a central figure as the fraudster, wheras Enron and the AI bullshit are institutional-level fraud requiring multiple companies to be complicit.
The new one is the most elegant, because they don’t have to break the law to make money on the grift. It also makes it the biggest by far because everyone wants in on the action.
This bursting of this bubble with be catastrophic. Maybe tens of trillions (there’s estimated to be about 125 trillion dollars in assets total worldwide).
And the .com bubble the hosing crisis, the banking collapse etc. etc. It all seems like the same scam being redone over and over. Never the same but always alike. And now it seems crime is legal (if you have the capital backing).
I really hope at some point in my life the fuck ups have some lasting consequences.
Usually it’s because someone needs the line on a chart to go up… no matter what.
American big tech hasn’t had a breakthrough technology equivalent to PCs, Smartphones, internet, etc. in ten years. Their latest forays into smart Watches, Smart glasses, VR that stupid virtual world Meta pushed… none of it has seen real adoption or growth.
They needed AI to be a hit to continue justifying their stock prices. The only way they can justify their stock prices now is if sales for AI become a 2 or trillion dollar industry. The only way you do that is by replacing humans. Human jobs go away and corps pay for AI tokens instead.
But it hasn’t panned out that way and so far and all you have are AI companies passing around the same dollar and booking it as revenue on their P&Ls. That will work for a little while but it isnt a industry.
I feel like everyone knows it’s a bubble but that people are expecting it to be like the Dot-com bubble where a small amount of companies succeed, hoping that they invest in those companies.
That’s pretty much what is driving the entire AI “boom”. It’s the Amazon effect.
Amazon, and Netflix to a certain extent as well as a number of other early internet startups, lost money for around 10 years before they started making shit tons. And everybody wants to be on that particular ride. So they are queueing up in every AI line they can find to hedge their bets on one of them being the one that survives and thrives. It doesn’t really occur to any of them that all of them could fail. Because “AI is the future”. Except there is no plan to monetize any of them. And so far every time one of them introduces a charge, people stop using it.
I feel like it’s just as likely that further breakthroughs cause costs to plummet to the point that any of the major AI companies will never even be capable of seeing a profit, paving the way for some yet unestablished or open model to reign supreme in the end.
Hopefully, anyway lol.
Let’s all short AI! Dimond hands!
How? The chatbot labs don’t have the unit economics to be taken public.
I guess you could short Nvidia and AMD.
I think nvidia is going to be one of the only sure winners here. They’re making hand over fist selling shovels during the gold rush.
It’s not everybody, though.
Let’s not forget that they’re literally stealing from anyone who ever made anything in our entire history.
The profits will come from reselling everything again as an add-on or inapp(ropriate) purchases
Fortune is part of this bubble. They helped create it by encouraging top management to ride it… And even here they still do “he said” reporting.
‘Profits are currently being funded by investors rather than earned from customers’
Which profits?
Ah, now I see the actual quote is: “[…] the AI boom’s profits are currently being funded by investors rather than earned from customers.” he means the profits of the shovel sellers, not AI profits.
nvidia is the only one that is making money from this?
Not just Nvidia.
For a start of who the shovel sellers might be, you can take a look at the list in his post that Torsten Slok used to calculate his margin numbers:
Note: Data as of 2Q 2026 and for OpenAI (1Q 2026 estimate from PitchBook) and Anthropic (2Q 2026 estimate from Financial Times). Averages are equal-weighted bucket averages of Energy & Grid (Constellation Energy, Vistra, NextEra Energy, Vertiv, Eaton, Arista Networks), Silicon & Equipment (Nvidia, AMD, Broadcom, Marvell, TSMC, SK Hynix, Samsung Electronics, Micron), Compute & Cloud (Super Micro, Dell Technologies, Foxconn, Equinix, Digital Realty, Amazon/AWS, Microsoft/Azure, Alphabet/Google Cloud, CoreWeave, Nebius) and Models & Applications (OpenAI, Anthropic). Sources: Bloomberg, PitchBook, Apollo Chief Economist
In short according to his research the people currently making a profit or at least positive margins, are the people who build chips and memory, who build energy infrastructure, and who build servers and data canters.
I would like to see the distinction between Servers (Super Micro, Dell Technologies, Foxconn) and Datacenters (Equinix, Digital Realty, Amazon/AWS, Microsoft/Azure, Alphabet/Google Cloud, CoreWeave, Nebius) but he seems to have put them together.
Also I heard from our local news that ABB (which is Swiss, Edit: apparently also Swedish) is profiting too, they belong in the power category.
And I wonder why Arista goes in the Energy category, I’d see them and Cisco, and Juniper, and Extreme, and Nokia in a separate network category.
Ponzi called, he wants his scheme back.
Well the jerk store called, and they’re running out of you!
Relax Costanza 😂
Well I slept with your wife!
Enron on the side asking all I had to do is wait for AI?
I’m willing to bet most of the funders are people using AI to replace workers and manipulate the masses. So really; they’re paying customers.
I didn’t go to business school, so take this with a grain of salt: doesn’t business solvency require you to charge more for your product than it cost you to produce it and bring it to market?
in theory, in practice you just need to be convincing enough to get greedier/dumber people to pay for your shit in the gamble they might get more in return later. (they like to call themselves “investors”, but most are just degenerate gamblers) and then cover your own ass if/when the house of cards tumbles.
building and popping bubbles, and fleecing as many people as possible throughout the whole thing, is what the US
financial marketscasino is literally designed to facilitatedeleted by creator
Powerful people are willing to pay a large price to manipulate the masses.
As7de from that, the plan is to lose money for years gobbling up all the data before there’s laws that actually get enforced to protect against it, and to get an entire generation of college kids and workers dependant on using AI, with companies relying heavily on its usage. This is happening to a huge degree right now with data entry, “vibe coders”, Law firms, and healthcare.
Those people won’t really be able to transition back or away from AI. They’ll be dependent on it. And AI will start charging them way more than they are now. And they’ll eat the costs and pay it.
So it’s stealing data while it’s free or nearly free, power to manipulate all the sheeple, create dependency.
The chatbot labs (and Oracle) quite literally don’t have years.
Everybody doesn’t get to win a race…
I seriously doubt you read that article if that was your only takeaway from it.
Nobody is winning the LLM lab race, all of their unit economics are garbage and every new subscriber actually makes that situation worse instead of better. Furthermore, I’m running a open 30 billion parameter LLM on my laptop via Ollama right now, so what the fuck do I need lyin’ Sam Altman for?
At this point I’m just here to watch Oracle crash and burn.
But it doesn’t work! You hear all sorts of anecdotes about a business who retrenched their workforce to replace with LLMs, everything comes crashing down and they have to rehire all their workforce at higher rates.
It would be interesting to see actual verified numbers. I am sure that a lot of useless people can be replaced with computers, but most of the time, all it requires is a very small sh script, it doesn’t need a monstrous data centre that uses more water and power than a human.
It doesn’t work for some people and some businesses and when that happens, it gets blasted all over the place because it makes for a good headline.
Tons of jobs have been replaced by ai, though. Especially customer service chats and online help and a LOT of game production and design art jobs and music and editing jobs.
Plus all the jobs where people are still there but heavily relying on ai.
At least in the tech industry that doesn’t apply if the investors believe that the technology or application is promising enough. Look at Uber’s example. They lost very large amounts of money for many years before turning a profit. They were able to do this because the investors kept sustaining them through their unprofitable years.
The same thing has been happening for years with the AI companies. I heard a while back that OpenAI wouldn’t turn a profit on their services even if they charged end users $100/month.
Good article, thanks! That provides better context for how the major AI companies’ finances compare with Uber’s, but I can’t think of a better example to use for a company that operated at massive losses for many years due to being propped up by investors. If you have a better example I could use, I would be glad to do it. Otherwise, if this question comes up again in the context of AI, I’ll be sure to include a link to that article.
There isn’t a better example. This level of FOMO bullshit hasn’t happened since the Dutch went bonkers over tulips.
Chatbot labs have the worst unit economics of any companies in recent memory. They only reason they still have doors open is VC money & that is now drying up. Unlike Uber, they will never become profitable. That’s why you are seeing the push to IPO so they can loot your retirement accounts via the stock market.
This level of FOMO bullshit hasn’t happened since the Dutch went bonkers over tulips.
Haha, now that’s a great comparison! I mentioned to a co-worker a couple weeks ago that we should get ready to invest in the next bubble coming up after this one pops, probably tulip bulbs (again).
Good points, thanks.
Uber is an example of government corruption leveraged to overthrow a market. The livery industry required special licensing, so only licensees were legally allowed to do it. Uber then started doing it without the $1 million per license overhead, bribed the government to ignore that, and naturally undercut their competitors that had huge government-imposed overheads that Uber didn’t. They were initially unprofitable because they had to compete for business, then became profitable when the government helped then destroy their competitors and structure their pay so as to have no minimum wage.
If you only look at modern examples, youre looking at a funhouse mirror reflection of real business operations. Running at a loss for a decade while you create a monopoly is the path to business success today, but this is not a functioning market. Real business success involves making money right out of the gate or at least right after. If your business only works if it is a near-monopoly, it is not a sound one.
I completely agree with your conclusion. Just to be clear, I was in no way saying that what Uber did and OpenAI does is OK. They both suck and should never be allowed to operate as they had and have. I don’t know if Uber was government corruption though, at least I had never heard it described as such. What I have heard is that they found loopholes in taxi laws around the world and employed a global army of lawyers exploit those loopholes.
At least in the USA, livery workers were required to have a “medallion”, which is essentially an onerously expensive license with the stated purpose of restricting competition to the point of profitability for those workers. There was no loophole, and it was ridgidly enforced.
Then, Uber came in and started violating that law so flagrantly that governments seemed stunned into inaction. The government corruption comes in where they then decided not to enforce those rules on Uber for some reason around the same time Uber aggressively contributed to political campaigns around the world.
I went down this line of discussion to point out that Uber is not a good example of organic business success that started out financially underwater. Their ethics or lack of them is an aside to the discussion, but agreed, they are unrepentant scumbags.
Got it, thanks. My memory of how Uber was able to gain a foothold was a little different, but my memory of it was patchy anyway. My recollection was that they essentially said that Uber was not a taxi service. They sold it as just regular people offering to “share” a ride with someone else to an area that they were going to anyway. I have no doubt that heavy political contributions were part of their arsenal in their attack on the taxi industry and regulations, so I agree that this was government corruption. But I also recall that they also had armies of lawyers fighting at every level of government around the world to ram through the idea that this wasn’t a taxi service and shouldn’t be subject to taxi regulations. Maybe my memory is more patchy than I thought, or perhaps both things were true - government corruption and a legal war allowed them flourish.
