So before we get into today’s spicey and ultimately divisive discussion topic, I just want to say to those of you who have been following this blog for awhile, thank you. I haven’t posted here very often since I transitioned most of my takes over to a more anonymous blog a couple years ago. Looking back a few years later, I have to admit that I regret doing that. This is primarily because I can’t help but to feel that adopting some anonymity has inherently reduced the range and impact of my diatribes on the tech industry and the world in general. This post marks the very first step that I’m taking to reverse that change.
But I’m not stopping there. I’ve decided that at least for the next few posts I make here, I’m going to also start posting video versions of them to a new YouTube channel. This is not something I’m committing to in the long term but rather me trying something new. Nowadays the general feeling that I’m getting is that people don’t enjoy reading. In addition their attention spans have been ruined by years of tweets and various other short form media. Paying attention for awhile is no longer en vogue. Anybody who knows me, knows that I feel that this is a symptom of a much deeper sickness running through the veins of our society. However my words, regardless of whatever form they are ultimately delivered in, will likely have more impact if I choose to adapt to this reality rather than spit in the face of it.
So be it.
On the flip side, one thing I’m adamantly refusing to adapt to is AI. Better yet let’s just stop calling it AI altogether because its completely artificial and not actually intelligent. It does not know things. It does not think. It does not reason. It cannot improve itself nor can it correct itself. It is not even an entity worth a sliver of the consideration that I would afford to a cockroach, a being I’m more likely to crush the life out of with my sandal than anything else.
In any event, in the five months since my last post where I bemoaned the continuing rise of LLMs, the situation has gotten way worse. Supply shortages for RAM and Storage have hit. Tech prices have skyrocketed. But fear not because I believe the tide is about to start turning. A couple of factors are working in the favor of the Anti-LLMers:
-
Of the front line so-called AI companies, SpaceX (after its xAI merger), OpenAI and Anthropic, the first two are moving towards their IPOs. This is excellent news for all of us because it means that the ability of these companies to keep their burn rates off the books is going to be severely hindered once they file their S-1 forms and even moreso once they go public. SpaceX has already done this out of the gate it was a veritable dumpster fire. Don’t believe me? See for yourself. In the meantime, here are a couple choice quotes:
SpaceX is targeting a $1.75 trillion valuation on listing day — larger than Microsoft, trailing only Apple and Nvidia — despite posting a $4.28 billion net loss in Q1 2026 alone and an accumulated deficit of $41.3 billion.
…
Despite the EBITDA profit, SpaceX posted a GAAP net loss of $4.94 billion for full-year 2025. Q1 2026 accelerated that trend with a $4.28 billion net loss in a single quarter. The accumulated deficit now sits at $41.3 billion.
The primary driver is artificial intelligence. SpaceX disclosed that xAI/AI operations posted losses exceeding $6 billion in 2025 and burned another $2.5 billion in Q1 2026. The company is reportedly evaluating in-house GPU manufacturing to bring those costs down.
OpenAI is reportedly gearing up for a September IPO and their S-1 numbers will likely be even worse because they don’t have the ability to try and cover up the losses by merging with a business that wasn’t a complete dumpster fire. Altman and his band of strung out AI addicts are just burning through stacks of cash with literally nothing to show for it. That of course is me ignoring the endless stream of half baked ideas that are better suited for a Sunday morning wake and bake session yet somehow were allowed to become a “business plan”.
What does this mean? Well it more or less it means that the market won’t be able to pretend that AI is a profitable venture anymore. Anthropic hasn’t announced any plans to IPO yet, but I’m sure it will happen in relatively short order. How do I know that? It’s because the Venture Capitalists backing all of these companies desperately need an out. They have plowed hundreds of billions of dollars directly into these companies and as of right now they don’t have even a nickel’s worth of profit to show for it. Once these companies go public, the Venture Capitalists will be on the path to dumping their shares onto unsuspecting retail and institutional investors which is pretty much what always happens right before a tech crash. The real rich people, always evade accountability and this time will likely be no different.
Make no mistake, the VCs know that these companies are shit. They know that they will not make a single dime worth of profit ever. That is why all of these IPOs will happen within the same few months. Because if any of them wait too long, the bubble will pop and the VCs behind that particular company won’t be able to recoup any of their losses. The only way they can do that is to dump their shares on the open market and the only way they can do that is the force these companies to go public.
This of course leads us very nicely to my second factor…
-
The remaining momentum around the AI hype train may save the squandered fortunes of the poor VCs staring down the barrel of only being able to afford a couple of McMansions and speedboats rather than full on compounds and city-sized yachts. However the real reason the bubble will pop is that none of these companies can afford to sell their LLM powered services at a loss. Public market investors will not tolerate this. In preparation for this new reality we are already starting to see prices rise on various fronts, perhaps the most notable example of which is Github Copilot:
TL;DR: Today, we are announcing that all GitHub Copilot plans will transition to usage-based billing on June 1, 2026.
Today, a quick chat question and a multi-hour autonomous coding session can cost the user the same amount. GitHub has absorbed much of the escalating inference cost behind that usage, but the current premium request model is no longer sustainable.
Usage-based billing fixes that. It better aligns pricing with actual usage, helps us maintain long-term service reliability, and reduces the need to gate heavy users.
This statement is a marvelous example of why so many of us hate Microsoft (who owns Github) and will likely never stop. It portrays Github as the long suffering provider of services and basically paints their heavy users as the source of all of the problems Github Copilot is having as of late. Keep in mind that one of the primary goals of modern day tech businesses is akin to that of a drug dealer out on the street corner: They get you hooked and after that they turn the screws and make you pay. This is the exact same playbook except the person fucking you over is wearing a suit and wields an expense account that would make even Pablo Escobar’s hippos blush in embarrassment.
Also, let me be REALLY CLEAR and state that the price increases for Github Copilot are insane. I know that because they released a tool which previews what your first bill in June will be by applying the new pricing to your usage from April and the results are absolutely staggering. You can see some of them in this reddit thread.
Mark my words, this is just the beginning. LLMs aren’t cheap to train or to operate. That is true in terms of hardware, power and other natural resource requirements. Once the horrendous negative margins around this business are fully exposed, the pressure is going to be on. That’s because once retail and institutional investors are left holding the bag, there is nobody left to pass the hot potato off to. But the joke will ultimately be on us because the people who created, crafted and curated these abominations of tech malfeasance will not be going broke. Because once the IPOs are done, they’ll have their out. They’ll be even richer… for a time. Once any restrictions around when they can dump their shares on the open market expires, I expect them to begin selling as much as they can as fast as they can.
The fix is in and the people running these businesses know the score. That’s why they are in such a rush to go public. They don’t have the liquidity or the credibility left to raise anymore money from private investors so they are now forced to go to the sucker of last resort: Us.
But hey lucky for them the Trump administration is running the show and they are not anything if not more overtly corrupt and willing to lie, cheat and steal so long as there is a profitable angle in it for them. So of course they are working feverishly to weaken the financial disclosure rules around IPOs and public companies in general:
Once upon a time, IPOs were the ultimate graduation ceremony for entrepreneurs. But over the years, regulations piled up like snowdrifts.
Quarterly reporting. Shareholder lawsuits. Endless disclosure requirements.
According to Paul Atkins, the current SEC Chairman and a Trump appointee, that’s a big part of the problem.
“Disclosure isn’t meant to be torture,” he said recently. “It’s meant to provide material information so investors know what they’re investing in.”
Atkins believes excessive red tape has turned the IPO process into a bureaucratic nightmare. That’s why he’s vowed to “make IPOs great again.”
I’ll tell you what, it’s a real shame that these pesky financial disclosure requirements are making it so hard for businesses to scam their investors and customers out of their money, isn’t it? Thank goodness Trump’s cult, MAGA, came along and propelled the king the of all morons directly into the highest office of the land. I for one cannot wait for IPOs to become great again. Maybe they’ll even go so far as to remove the requirements for companies to at least mention the GAAP numbers when they provide their own non-GAAP versions of them because having to provide even the slightest semblance of reality in their reporting is obviously hindering their path to increased profitability, am I right?
Sarcasm aside, ultimately I don’t think these changes will matter. That’s because these AI companies are losing so much money, it can’t really be effectively hidden. There is no home for a hundred billion dollar hole in any balance sheet, much less the balance sheet of companies that simply aren’t making THAT much money to begin with.
In any event, prices for these services will rise dramatically because they simply have no choice now. Even if Anthropic wasn’t originally planning on an IPO this year, I’m betting they’ll have one anyway. Last one to the dance ends up being a lonely fellow and I promise you that Anthropic’s VCs absolutely do not want to be the stags at this party.
So where does that leave us? Well I believe we are on the precipice of the so-called AI bubble popping and I believe I’ve made a strong case for that in this post. I believe that the IPOs will only serve to accelerate the endgame for a business model that never had any business being taken seriously. Normally I’d follow all of that by saying that I believe that this will also kick off a serious recession, but given that Trump and his wrecking ball administration has broken virtually everything else I’m officially upgrading my prediction to a serious depression.
Consumers like you and I are strained. Every week we feel the pain at the pump, at the grocery store and in every single purchase we make. Our household has already cut back on a number of things. For example, I only pay for a handful of streaming services now. Three in particular and one of them, HBO Max (or whatever the fuck its called now), is included with my AT&T Fiber Internet service. The other two are Apple TV+ and YouTube Premium.
I speed ran every pragmatic tech upgrade I could over the last year to get ahead of this. I have been stockpiling non-perishable food in the event that the fertilizer shortage created by Trump’s ill-advised war in Iran creates a food shortage and possible famine a few months from now. We have even been dealing with maintenance issues in the house that we put off just because we don’t know if we’ll be able to afford dealing with them six months or a year from now.
The immediate future is not looking very rosy, but I want to be clear: I believe that collectively we will come through this. We as a species are better than this. Right now, at least in this country, what you are seeing is us declining towards our worst imaginable state. But the decline will only continue as long as we have the stomach to sit back and allow it to. For this to end and for the sins of those driving it to be properly adjudicated we will need to eventually seize the reins of power from those currently driving our collective wellbeing into the ground.
There is no other way.
Together we can be better, but only if we choose to be. Together we can stop this, but only if we choose to exercise our inherent right to do so. Whether it be from AI, Trump, MAGA, Fascism or the new age Robber Barons that see themselves as the ordained puppet masters of this our new gilded age, we can change this course and I think given enough aggravation, pain and suffering we will do exactly that. It’s only a matter of time.
May fate treat all of you kindly until the next time we meet.