Unpopular Opinion

I disagree. This is so much like the dot com era. I mean much like dotcom the web did not disapear and ai won't disapear when ai busts. Bust just means recognition of the lack of return and a massive crash in the stocks.

Tell me you're not an economist without telling me you're not an economist.

I think there will be a time when it costs too much to keep a data center safe and operational.

I went to a town meeting where someone is pushing to get one of these data centers built. People are not happy about this. Nobody supports it. And when it gets built anyway, there will be a bunch of people who suddenly can't afford their electric or water bills, and there's a very definite and universally hated reason why those bills are suddenly so high. That reason has a physical location and an address, along with a lot of vulnerable infrastructure.

I went to that meeting expecting to be the only person against the data center. I was wrong. People are mad.

The internet bubble was not about internet never being successful. It was a financial investment bubble based on overinvestment/inflated share values far beyond the profit extraction power of those companies. Pets.com is a posterchild of business model absurdity. Worldcom and CISCO also crashed due to absurd hype overinvestment levels. The housing/GFC bubble was not the end of housing or banking.

There is a similar bubble in AI investment, where LLM models have zero competitive moat, and cannot ever make money, where datacenter competition is rising faster than demand, and also cannot make money based on economic value, because of competitive pressure.

The only hope for AI surviving a financial bubble is the US government buying all the excess datacenter compute for surveillance and military skynet. This is extremely likely if only to ensure Zionist supremacist rule over the US, but also because the US is desperate to pick one last sector of competition with China, before losing at everything. Zionist supremacist rule over America political establishment unanimity, is matched by warmongering supremacy ambitions against China led multipolarism/global south.

The only alternative to an AI bubble, is an everything else collapse, and US bankruptcy Skynet led genocide of Americans for US establishment supremacy objectives instead of "machine supremacy". There can never be any threat to Zionist/warmongering supremacism in America, and you will all die before any shift to humanism/freedom dividends/UBI occurs.

2. Bubbles don’t HAVE to burst anymore Tesla hasn’t put out a successful new product in 20 years, and it continues to barrel right along, with its useless hack CEO hanging on as the richest person in the world. The old rules do not apply, and the sooner we acknowledge it, the sooner we can prepare for the new normal.

But it sells a useful product?? It sells an electric car and sells the only one that governments seem to be aware exists, since all the charging infrastructure is for Teslas.

Collectively burning tens of billions per month for negative ROI. The quarters will start turning red, investors sell, prices tank, covenants are violated, margins are called. Etc etc. plus throw in a few more Salesforce stories where they fire people, ai screws up and they publicly have to rehire and admit it was a mistake or moltbook style vibecoded disasters.

Just like .com bubble, it doesn't kill the tech, just the people that overinvested in infra for something that is only growing organically and will soon hit a natural plateau in improvements.

This was clearly written by an AI chatbot and is complete bullshit.

Barring a revolution, we are on an inexorable course to the most awful tech dystopia imaginable.

Sounds like a call to action....

So there was this Roman empire....

I agree that company fundamentals don't matter, and that the goal right now seems to be techno-feudalism.

Retail is ~40% of the stock market, and retirement is ~20%. I think it will pop whenever the next recession starts. If enough people lose their jobs, and start selling their stocks and dipping into their retirement to pay bills, these stocks will tank. It's my theory that these stocks are held up by all the people passively investing in index funds following the SP500 and NASDAQ. During any small sell-off by active investors, a lot of people's retirement plans will eventually automatically just buy and slowly bring it back up.

A scare, widely reported by media, could also cause people to take enough notice where a lot of people manually adjust their plans. But, yeah, the media is pretty captured now, and will avoid reporting on stuff that could cause scares.

WHO is going to allow the bubble to burst? The investors who would lose everything? The SEC that would collapse the entire economy by not turning a blind eye? The captured politicians that are being paid billions to be complicit?

I don't think you understand how bubbles burst. They don't burst because someone allows them to burst, they burst because someone runs out of money. When investors back off, money runs out. OpenAI is the weak link in this entire circle jerk, and Nvidia and Microslop has just backed down on their hundred billion dollar deals with them.

Sorry chief. Just not convinced of your argument at all.

Not adopting the AI paradigm is going to become increasingly costly

Why do dev machines cost $7k now ? Can someone who works for a large software development company confirm ?

Honestly I'd have thought things are going the other way. My laptop that I purchased for $500 several years ago is a great daily driver. I dabble in development, some is local, but LLM stuff is offloaded to an inference API, or a bare metal server which I rent.

I understand that sophisticated development companies aren't buying second hand laptops, but I don't think there's a sudden imperative to buy everyone $7k dev machines every year.

Bubbles don’t HAVE to burst anymore

You haven't really offered much to support this assertion.

I can assure you that any bubble will burst if there's an interruption in shareholder sentiment.

Things aren't going so well for Tesla lately. It's interesting that their CEO is sinking many millions into conservative political campaigns.

NOBODY who is responsible for enforcing anything like responsible economic activity will EVER allow the bubble to burst

That's not how bubbles are maintained though.

Right now, everyone's pension fund has invested in the "magnificent 7" because frankly, no one can afford not to.

Everyone knows the shares in these companies are overvalued, but no one knows by how much, and no one knows when the correction is going to come.

Its good for an unpopular opinion.

Counter point: a boat got turned a bit sideways in a canal and the entire financial system almost came crashing down and there were no bunnies for Easter.

My point is that very small, unknown things can cause huge interruptions to global functioning, and almost everything about the current bubble depends on a very stable global order, where specifically, the west, has access to the chips, when the west does not manufacture them in spite of whatever agreements are on paper.

A small shock to the supply chain and it all falls down.

The dotcom bubble burst. We still use the internet

The same greed that created the bubble, will burst it. Some are buying stocks to increase their wealth, others are shorting stocks to increase their wealth. The ones buying the stocks will soon run out of money to invest into a technology that doesn't return any profits.

I think a lot of the reasons you have pointed out are why the bubble hasn’t burst and won’t soon. It took 9 years after the dismantling of the Glass Steagall act for the Great Recession. The economy just has a lot more momentum than people think it does. The people near the top do not have enough power and control over the economy to stop everything forever though.

Small note:

negative feedback loop

That's a positive feedback loop, btw.
Reminder that pos. vs neg. is not about good vs bad, it's about increase vs decrease. Positive = increasing effects, negative = decreasing.

The dot com bubble burst but we still have more interwebs then ever.

AI can be both a bubble that pops and an inevitably large part of the future - the two aren't mutually exclusive.

Your argument is invalid right from point 1. Machines are increasing in price because hardware vendors are jumping into the bubble hoping to inflate it before it bursts.

If we are going to go back to costs, both are equally expensive. Employees take time and money to train, but they are flexible and can create new things. They can react appropriately in an emergency, and they have context that helps them avoid making poor decisions.

AI takes development effort to implement which ends up going to the employees anyway. AI companies are going to need to recoup rising hardware costs somehow, and that means the price to license AI is going to go up as well. The AI may or may not do what you want and will continue to be poisoned until a PAID employee tells it otherwise. The damage bad AI can cost is several degrees of magnitude higher than a bad employee. At least with a real person you can recoup some of that loss or have legal liability.

Like all emerging tech, it starts out really bad and grows as the developers make enhancements. AI may get there one day, but it is currently way too damn early to crash our economy over. The tech needed to mature more, and as is custom with big investors they are not easily willing to admit they were wrong and pull out. The bubble will pop - people are not happy with having immature technology shoved down their throats, and they are not happy with subpar tech taking their jobs.

regular people are NOT going to spend that for a PC. Which means most of the software that’s going to be developed will target iOS and Android… and that’s it.

Were you in a coma for the last 10 years? Because it seems like you missed the part where this already came to pass many years ago.

The bubble won't burst so long as there is more money being printed and companies can just inflate their way out of debt.

The bubble will burst because investors eventually reach a point where they run out of money to invest in a product that does not return profit.

OpenAI burns more money than it makes in revenue. And it's not going to get any better because LLMs are basically at the limit of how much they can be improved after already training on basically the entirety of recorded human knowledge.

There also isn't evidence that AI improves productivity enough to make it worth the cost, especially not the true cost of AI. AI services are artificially cheap for users because AI companies will eat the majority of cost with money they get from investors, but if they started charging something like 10x the price to cover the true costs of running the AI, you'd be hard-pressed to find any company that can justify continuing to use the product.

I think it won't pop either, but I'm hopeful that it will, and take a lot of rich people with it. One can dream, right?

It's an interesting point of view bud I always cringe when I read:

developers highly paid

Not in France where I live. We are not highly paid, have no regular raises, we can be fired for any reason, and we do unpaid overtime constantly.

A development machine that might have cost $2000 a couple years ago is well on the way to $6000-$7000 in the near future.

Trust me bro.

t’s not going to pop. Barring a revolution, we are on an inexorable course to the most awful tech dystopia imaginable.

In fact, right now, we see lots of companies pushing away from AI because it is unreliable and problematic. But the alternative costs from the other side are creeping up.

You say this and yet your arguments proceed on the assumption AI can actually do anything useful

But the alternative costs from the other side are creeping up. If you are a company looking to hire developers to write software, you need to provide development machines to those developers. A development machine that might have cost $2000 a couple years ago is well on the way to $6000-$7000 in the near future.

This isn't true because development doesn't have to be on a max powered system. Myself, I work on AS400 stuff. The compiler is on the remote machine. The database is on the remote machine. The JVM is on the remote machine. All I need is an editor and the ability to SSH into the machine, that's all that's required for development. SQL queries? I do those in a web tab, the query is ran on the remote machine.

I could easily write all the code, debugging, SQL queries, ORM, the Java stuff, node.js stuff (Yes, AS400 has node.js and it works pretty well with COBOL objects), and so on, on a Raspberry Pi if I needed to. And this might surprise people and then there will be people who this won't surprise. Because waaaaaaaaaaaayyy back, that's how it actually worked. You had a terminal and the machine you wrote code for and debugged and what not was in the basement and you were talking to it via some twinax. We've done way more development on a machine nowhere near us than we have on our own machines historically speaking.

As for the mobile stuff. That'll figure itself out. Or it won't and we just have wrappers called apps, that are just fullscreen web page. But again, the AS400, we had a product from Profound that basically wrapped a web app into a "native" app and the product from Profound handled all the stuff like taking a picture and putting on the IFS for an RPG program to pick up.

There's all kinds of ways around this notion that devs won't have beefy laptops, that's literally the way we used to do it for nearly forty years.

Tesla hasn't put out a successful new product in 20 years, and it continues to barrel right along, with its useless hack CEO hanging on as the richest person in the world.

This conflates a ton of unrelated things. Tesla has been unseated by BYD, but the US Government still is hung up about Chinese car makers enter the US market. Musk is the richest person because his wealth is largely predicated on the US economy, which because of the previous international cooperation stood as the leader. Everyday we are witnessing the US becoming weaker on the global stage, which means that his wealth translates less and less to things outside of the US. The US Stock Market is only a big thing because the US and the dollar is the thing so many things are pegged to. When that ceases being the case, the US Stock Market loses value in an international sense and since Musk's wealth it largely tied to that, so too does his wealth go down.

But the success of Tesla, isn't actually success, it's story of the incredibly sorry state US automakers are in that they can't provide a solid alternative. But don't get confused. Tesla sales are slowing drastically. This why we are seeing consolidation of Twitter, Tesla, and SpaceX. They're being consolidated because they're hitting rough patches standing alone.

So please don't confuse the odd situation of Tesla with Bubbles don't HAVE to burst anymore. Those are not correct conclusions here. Ford, GM, and the rest of US automakers are so down BAD at the moment that Tesla is able to shine. That's really the only thing keeping them floating, US car makers are jokes at this point.

NOBODY who is responsible for enforcing anything like responsible economic activity will EVER allow the bubble to burst

Greed. That's what you are speaking of. But greedy people come in all kinds of flavors. And there's no shortage of people who short the market and make mad cash doing so. Greed is universal, but the "upside" (I guess we'll call it that) is that the system allows greedy people to bet against the system. There are people putting money into this whole thing crashing down and the bigger the fall the bigger the reward. There were a ton of people who make billions when the housing market crashed. People watched 9/11 and were betting that the market would collapse in response. Don't underestimate people's greed. There are people who would bet money on innocent people getting shot if the odds were good.

Now all that said, there's a difference between this AI bubble and the technology. Like if the bubble pops, AI will still be a thing, the bubble is not AI itself, it's how we're developing AI at the breakneck speed we're going at.

Upvote not only because it's an unpopular opinion, but also flat-out incorrect/exaggerated.