Matías Potel Feola
LightNews LightNews
banner
harrypotel.bsky.social
Matías Potel Feola
@harrypotel.bsky.social
120 followers 690 following 1 posts
Posts Replies Media Videos
Reposted by Matías Potel Feola
edzitron.com
I cannot see how the AI bubble continues past 2027, and see many signs that it'll burst in 2026. OpenAI is running out of cash, plays silly games with user numbers, and has a future contingent on a very, very tough for-profit conversion. Things look grim.
www.wheresyoured.at/ai-bubble-20...
What Happens Next? This has now been the longest thing I've ever written, and the most I've ever written in the space of a week, but I wanted to try and estimate what might happen next as quickly as possible. NVIDIA's earnings will be groundhog day for AI. If Jensen doesn't see his shadow, earnings will beat expectations and growth will continue, and the markets will get a little hopium for the next three months. But I don't think, even with successful earnings, that this narrative goes away. Big tech is not showing any return on investment for AI, and OpenAI blew the launch of GPT-5 so hard that it's unclear where this company goes from here. Both OpenAI and Anthropic are dependent on endless resources at a time when the markets are hinting that they're going to need to see those resources actually do something. I will also be honest: I have no idea how OpenAI converts, as I have no idea how SoftBank gives it the money if it does so. I also have no idea how OpenAI or Anthropic go public, and they must go public, because they are too big to acquire, and going public will require them to show all of their dirty, ugly financial statements. Even if OpenAI raises its $20 billion, it needs to raise another $10 billion or more in 2026, and Anthropic is in a similar position. How much more venture capital is left to prop up these capital-engorged failsons? How long will Silicon Valley tolerate two giant losers that love to lose money?This entire narrative — this entire industry — is based on vibes, and there are enough signs that something weird and dodgy is going on that I believe someone is misstating how much they make, how much they lose, or a little of both.
August 27, 2025 at 3:04 PM
Reposted by Matías Potel Feola
edzitron.com
OpenAI must convert to a for-profit by end of year or SoftBank won't give them $20bn. There's a scenario where Microsoft - who effectively owns their IP and research - lets oAI die and scoops up the assets.

Even if they don't, OpenAI doesn't have enough time.
www.wheresyoured.at/ai-bubble-20...
Chaos Bet: Microsoft Kills OpenAI By Blocking Its Non-Profit Conversion As I wrote a few months ago, I think that Microsoft may want OpenAI to collapse. The Information reported in June that, when funding OpenAI's 2023 round, top Microsoft executives believed OpenAI would fail, and the nature of their agreement — which gives Microsoft full rights to sell OpenAI's models and access to their intellectual property and research — means that OpenAI's failure would allow it to effectively take all of OpenAI's stuff. Though nothing is certain, there is a scenario where — as reported back in June — Microsoft simply walks away from negotiations. I imagine if this seems likely that OpenAI will continue its belligerent shadow PR campaign where it continues to accuse Microsoft of things like antitrust violations, and if Altman begins taking a public stance on the matter, it means that the negotiations have fully broken down and it's time for the fun to begin. While both parties are still negotiating, it's been a while since we heard an update. OpenAI wants Microsoft to reduce its cut of OpenAI's revenue, give up IP rights, and a bunch of other crap that Microsoft doesn't have to agree to. OpenAI's logic, I imagine, is that Microsoft would benefit from an OpenAI IPO. Remember, Microsoft has a lot of insight into the state of OpenAI, both in terms of its intellectual property and its physical infrastructure (which Microsoft runs almost the entirety of). Microsoft likely knows how busy its servers are, and can probably extrapolate revenues as a result, if it doesn’t already know them. Surely if things were going amazingly Microsoft would work with OpenAI to make this deal happen? Wouldn't it want 33% of the world's biggest, most hugest, most important company ever? I get the sense that this is not really a "negotiation" so much as it's Microsoft saying it’s happy with the terms of the current deal, knowing that its say-so is the only way to get OpenAI to resolve its non-pro… Even If Microsoft Agrees, OpenAI Does Not Have Enough Time To Convert To A For-Profit By The End Of The Year People are so fuckin' cavalier about this! Even if Microsoft gave its blessing tomorrow, OpenAI would have three or four months to pull off something that has literally never happened before, and have to get past both the Delaware and California District Attorneys. This is a complex, multi-faceted legal challenge with little or no precedent, and while I'm no legal or non-profit expert, I see no way it gets this done by December 2025. SoftBank has also not been entirely clear on whether the current remedies to the for-profit situation actually work. As I wrote up in July, SoftBank isn't exactly clear (thanks in part to weasel wording and translation from Japanese to English) as to whether OpenAI's current reformation will actually fulfil its requirements: Recently, OpenAI updated their blog. And also the media has been covered, and there was some misunderstanding. See, like you said, that the media is saying giving up on becoming the profit organization. I don't think that's the correct understanding. The shareholder of the – or the holding company is the nonprofit. And underneath the company which does the business, but more specifically, nothing has been changed at this moment. They just show the direction of the structure. That's what's been announced. So at this moment, nothing has been really happened or changed, just to discuss the direction. And for us, I don't think that that's a wrong direction, a bad direction.That's something that we've been expected. And our investment is in the business vehicle, and this is – hopefully that they'll be able to grow. Then that I believe that's a good thing for our investments. When asked about whether “the event” (referring to OpenAI not converting and thus SoftBank cutting its round in half), SoftBank's CFO said the following: No. Event is not happening now. So if such event takes place by the end…
August 27, 2025 at 3:04 PM
Reposted by Matías Potel Feola
edzitron.com
One scenario I've extrapolated - a chaos bet - is that one AI company is misrepresenting its revenues or costs due to a lack of due diligence on the VC side. It doesn't make sense how much money AI companies are burning, and I see no way many of them survive.
www.wheresyoured.at/ai-bubble-20...
Costs Are A Brewing Scandal In Generative AI — Another Chaos Bet Never in my career have I ever seen an industry given more grace than generative AI, with reporters gladly using figures like "$50 million annualized revenue" to make a company like You.com seem worth $1.4 billion, despite that working out to $4.16 million, or the revenue of four Dunkin' Donuts franchises. To be clear, I am not talking about any specific startup when I say this, but something doesn't make sense about the burn rates on these companies, or at the very least, the only way it makes sense is if they're apocalyptic. Why does Cursor need $900 million, and then needs to raise more funding? Where is the billion-plus dollars of funding that went to Perplexity, and why does it need more of it already? No, really, let's look at Perplexity. In July, it raised $100 million in financing after, according to Bloomberg, trying to raise $1 billion in March and, I assume, failing to do so. This leads me to two questions: Why couldn't Perplexity raise $1 billion in March, and why did it raise a tenth of that in July? What the fuck does Perplexity need this money for? I believe that beneath the surface of these companies are incredibly bad costs, so much worse than anybody has accounted for. While Perplexity's costs look "bad" based on The Information's reporting, with a net loss of $68 million, it's more than likely based on what it’s raising that costs are both much higher and worse than the reporting suggests. Remember, these companies (I am not saying this directly about any firm) have been allowed to run fast and loose with their costs, which more than likely means that somebody is undercounting or representing them in a more positive light than they should. For example, Anthropic's gross profit margins don't include the costs of training its models, and Perplexity doesn't include the cost of its free customers in its gross profit margins. Playing silly buggers with accounting only sugge…
August 27, 2025 at 3:04 PM
Reposted by Matías Potel Feola
edzitron.com
AI startups have ridiculous valuations, terrible burn rates and no chance to IPO or get acquired. Worse still, VC dry powder - available capital to invest - will run out by the end of next year if they keep investing at this rate. At some point VCs have to stop.
www.wheresyoured.at/ai-bubble-20...
The End Of AI Startup Funding As I mentioned previously, The Information wrote last week that the "dry powder" (available capital to invest in totality) has dropped, and venture capitalist Jon Sakoda of Decibel Partners believes that "...if VCs keep investing at today's clip, the industry would run out of money in six quarters." Sakoda adds that if you remove OpenAI and Anthropic, the money would continue through 2028… but let's be honest, those are the companies that are going to be raising the money, along with other bloated, overvalued AI firms. The Information estimates there's around $221 billion left of dry powder in the hands of venture capitalists, and Sakoda believes that this will drop to $164 billion by the end of 2025. The way VCs have gotten around their lack of capital is to use Special Purpose Vehicles — entities that can raise a chunk of money from a pool of investors — to pull together cash that would otherwise come from their Limited Partners and funds, allowing them to boost big bets without affecting their investment strategy (or pissing off LPs). Yet Anthropic, which is currently raising as much as $10 billion, has made it clear it will no longer accept SPVs, citing that their creators often have no connection to the company, taking fees for investors to participate — something that's true in some cases but a peculiar way to stymie the ability to raise capital. Nevertheless, venture capital is going to eventually run dry at this rate, because (as I discussed recently) the valuations of AI companies are such that they're too expensive to acquire, and their low-revenue high-burnrate businesses are tough-to-impossible to take public. Venture capitalists are simply not seeing returns on AI investments, which in turn will make Limited Partners far less likely to want to continue backing these deals, especially if the market sentiment shifts against AI. Right now, there are no profitable generative AI companies. Yet Cursor/Anysphere is valued a…
August 27, 2025 at 3:04 PM
Reposted by Matías Potel Feola
edzitron.com
NVIDIA is a load-bearing stock, 8% of our stock market, 19% of the magnificent 7, and the only company making *any* profit from AI. Their growth WILL slow, and once it does, the market will lose confidence in the AI growth story.

Big tech will follow suit.

www.wheresyoured.at/ai-bubble-20...
Nevertheless, NVIDIA Will Accelerate The Collapse If Its Stock Falters As I've said, NVIDIA sells physical things — GPUs — and those GPU sales are the only real barometer of the "health" of the AI bubble, as nobody else is expressing revenue (and certainly not profits). When the DeepSeek situation happened earlier in the year, NVIDIA dragged the rest of the Magnificent Seven (and stock market) down with it, because NVIDIA's revenue is basically the only tangible value in the AI trade. If anything happens to NVIDIA's value — a bad earnings, a DeepSeek-esque panic in the market (such as the launch of its R2 model, which is expected late August), or even just a dramatic vibe shift against the AI trade — it will drag every stock down with it. NVIDIA's current price-to-earnings ratio — a standard way of measuring how cheap or expensive a stock is by measuring their price-per-share versus their revenue-per-share — is around 50 as I write this, which suggests that it's likely overvalued (I'm not a stock analyst, to be clear), and the market could decide, on a dime, that it needs a haircut. Alternatively, investors could believe that NVIDIA has room to grow from here, which would mean in about a year it’d be selling $65 billion in GPUs a quarter, and upwards of $80 billion to $100 billion a year after that. This would require the Magnificent Seven — who make up 42% of NVIDIA's revenue — to buy more chips, every quarter, without fail, possibly until eternity. I don't know if that's going to happen. And remember: without NVIDIA, the only other support for the AI bubble is vibes. So, in simple terms, NVIDIA's earnings are the next big turning point for the AI bubble. If the market accepts its earnings report, we're in this for another three months minimum. If the market starts getting pissy, things will begin to accelerate, because all of the companies in the markets who have been saying they love AI will have to explain why.
August 27, 2025 at 3:04 PM
Reposted by Matías Potel Feola
edzitron.com
The common thread of all of my pale horses is they're impossible to ignore - unquestionable events that show that there *is* no AI revolution. In this piece I lay out every critical event that will make the bubble burst.
www.wheresyoured.at/ai-bubble-2027/
The common thread through all of these points is that they are predominantly impossible to ignore. So far this bubble has inflated because the problems with AI — such as "it doesn't make any money" and "burns billions of dollars" — have been dismissed until very recently as the necessary costs of the beautiful AI revolution. Now that things have begun to unravel, the intensity of criticism will increase gradually, rather than in one big movement that makes everyone say "we hate AI." And it isn't just because of the money. CEOs like Tobias Lutke of Shopify have oriented their companies’ entire culture around AI, demanding in his case that "employees must demonstrate why AI cannot be used before requesting additional resources." Generative AI is, on some level, a kind of dunce detector — its flimsy and vague use cases having enough juice to impress the clueless Business Idiots who don't really engage with the production that makes their companies money. The specious, empty hype of Large Language Models — driven by a tech and business media that has given up on trying to understand them — symbolizes a kind of magic to these empty-headed goobers, and unwinding their "AI-first" cultures will be difficult...right up until the first guy does it, at which point everybody will follow. AI has taken such a hold on our markets because it's symbolic of a few things: Executives and do-nothing middle managers' ability to control and suppress labor by suggesting a tool exists that can replace it. The future of automation, even though Large Language Models are absolutely terrible at it. The validity of the "ideas men" that run large parts of our society, who believe that their superior brains make them "above" labor somehow. In any case, I am going to try and write the things that I think will happen, in detail. I'll go into more conditions in this piece, and as discussed, I'm going to make some informed guesses, extrapolations, and give my thoughts about how things …
August 27, 2025 at 3:04 PM
Reposted by Matías Potel Feola
edzitron.com
The AI bubble is bursting, and the things that will accelerate the collapse are NVIDIA's growth slowing, the death of AI startup funding, the collapse of a major AI startup, big tech turning on AI, and a collapse of neoclouds like CoreWeave.
www.wheresyoured.at/ai-bubble-20...
NVIDIA's Growth Slows: As discussed in the Hater's Guide To The AI Bubble, NVIDIA is the weak point in the Magnificent Seven (which collectively account for 35% of the value of the US stock market) specifically because of its success (19% of the Magnificent Seven's value) and value, which is driven entirely by its ability to sell more and more GPUs every single quarter. It is inevitable that its growth slows, and once it does, the AI story crumbles with it. Perhaps it has a down quarter then an up quarter — but that's three long months to make the markets happy. I will add that there's a chance the bubble reinflates a touch if NVIDIA crushes earnings. Or maybe the market doesn't care? We'll find out at the end of next week. AI Funding Will Start Drying Up, And AI Companies Are The Most Funding-Dependent Companies of All Time: As The Information wrote last week, the "dry powder" (available capital to invest in totality) has dropped, and venture capitalist Jon Sakoda of Decibel Partners believes that "...if VCs keep investing at today's clip, the industry would run out of money in six quarters." Removing OpenAI and Anthropic, the party will continue until 2028...but we all know those are the companies that are getting the money. One of the major AI companies will collapse: OpenAI and Anthropic both burn billions of dollars a year, and have shown no interest in stopping doing so, with Altman doing his best Lord Farquaad impression and saying that OpenAI is "willing to run at a loss" as long as it takes to get...somewhere. For the bubble to burst, one of these companies has to die, and I will explain how this might happen. In both of these cases, going public will be nigh-on impossible, and even if successful will expose what I believe to be their rotten economics. Big Tech Will Turn On AI: Meta's AI hiring freeze isn't enough: Meta, Google, Amazon or Microsoft needs to bring an end to their capex burn, and they need to be definitive that it's A) happening and B) that it's happening, at least in part, because they have "built enough" or "exceeded the opportunity in AI." These companies are barely making $35 billion in revenue from AI in 2025, and I doubt that revenue is increasing. Another part of this: The Markets Make Big Tech Put Up Or Shut Up: At no point have the markets really interrogated the revenue from AI. I can imagine some weird fucking games happening here. Microsoft, the only member of the Magnificent Seven outside of NVIDIA willing to talk about AI revenue, stopped reporting AI revenue in January, when it was at $13 billion "annualized" (so a little over $1 billion in revenue a month). One has to wonder if this means revenue is flat or falling, as annualized is monthx12. I can also see a scenario where these companies start putting out obtuse "AI-enabled" revenue stats. I do not think this works, and if it does, it only puts off the inevitable, because they cannot avoid the capex crunch that's coming. This particular scenario is one that will happen in pieces. AI Startups Will Start Dying: As I've discussed previously, AI companies are currently raising at suicidal valuations that make it impossible for them to ever sell or go public. CoreWeave Will Die: AI data center developer CoreWeave is a time bomb, burdened with, to quote analyst Gil Luria of DA Davidson, "deteriorating operating income." It’s also a critical partner to OpenAI, providing compute as part of an $11.9 billion five-year-long contract, despite the fact that it's unclear if it has even come close to finishing its data center development in Denton, Texas. This has to be done by October, in part because that's when OpenAI is due to pay Coreweave, and in part because that's when it has to start paying back its massive, multi-billion …
August 27, 2025 at 3:04 PM
Reposted by Matías Potel Feola
edzitron.com
I will also add that anyone suddenly veering super hard anti-AI right now is suspicious too, and you're gonna see way more of them. Can't wait to see how legacy media tries to backpedal, especially as I've been reading every single god damn generative AI piece that's run for two years
August 19, 2025 at 8:50 PM