I really get the feeling that the economy is being artificially propped up at the moment and the key event I suspect they're being propped up for is the midterm elections. Post election, I feel like the cracks will start to show and we'll start to get a correction which will lead to a crash, similar in size to 2008.
That's just my feeling and it isn't necessarily backed on too much but significant amounts of resources have been poured into "AI" and I really don't feel the payoffs are going to pan out. Once that becomes the mainstream view, everything built around it crumbles.
The AI bubble will burst but "post-midterm" is relative. Based on previous tech bubbles it's probably still going to last for a few years.
Half wrong. It is currently "propped up" but that's because the US has a shit ton of debt to refinance in '26 (8 Trillion). The squeeze will follow that because then there's no need to flood the system to manipulate interest rates until next time.
AI is current tech darling but it's independant of the current circumstances. Once it plateaus they'll focus on quantum processing again to bypass limitations and throw money at orbital data centre infrastructure (orbital payload delivery, solar, battery, satellite to terrestrial network solutions). They'll probably be the main focus for the next 8-10 years. Data centres will be first use and the test case for further development, so the infrastructure around it is the better investment over the centres themselves, like AI, if we're looking very long term.
100% You can practically never go wrong assuming that we'll need more energy tomorrow than we do today.
Someone will probably eventually come along and tell them to stop using open loop cooling, there will be some money to be made there too even if they'll use in-house solutions in the end.
You say that, but right now everyone is jerking themselves raw over $60 oil. These data centers need a miniature sun to power, yet energy acts like we're in a recession. Where's the building, the drilling, the flow of commerce?
Wait... the data centers have frickin cooling ponds and towers now or something? Good grief...
Why would that be a bad thing compared to them just dumping their water into the sewer system?
The traditional way to water cool electronics is with a closed loop and a radiator, like a car, or a large HVAC system. It doesn't evaporate or get dumped in a sewer, it gets continually reused for a year or more, just like a car. As far as I know, all pre crypto/AI boom civilian data centers were either air cooled(overwhelming the most common, sometimes with a central system of high static pressure fans), HVAC cooled, or closed loop water cooled(and that was darn rare before crypto).
An open loop would be more like a nuclear plant. Dumping it in the sewer would also be an even less efficient open loop.
What are they going to do for cooling?
radiation
Radiation is the least efficient form of heat transfer. Those orbital data centers will cook themselves to death with their own heat, to say nothing of them receiving the full brunt of the sun's rays.
As an example- the ISS is capable of radiating 70kW of heat :
https://en.wikipedia.org/wiki/External_Active_Thermal_Control_System
Meanwhile, datacenters generate enough heat that they're measured in mW.
https://www.achrnews.com/articles/146987-data-centers-get-larger-hotter-making-them-attractive-sources-of-heat
ChatGPT suggests one launch of starship could take that much cooling into space :
ISS cooling system (radiators): how many watts can it radiate? • The ISS External Active Thermal Control System (EATCS) (the big ammonia loops feeding the external radiators) is capable of rejecting up to ~70 kW total to space (≈ 70,000 watts). 
ISS radiators: how much does it weigh? • A single Heat Rejection Subsystem “Radiator ORU” (the deployable 8-panel radiator unit) is about 1,082–1,123 kg (≈ 2,380–2,475 lb) depending on the cited source/rounding.
How many kg can Elon’s Starship carry? • SpaceX’s current public spec is ~100–150 metric tons to orbit in a fully reusable configuration, i.e. ~100,000–150,000 kg (typically stated for LEO payload).
Shh, don’t bring the laws of physics into an emotional argument!
I don't think midterms has anything to do with it. The USA is already at wartime/major economic problem level of deficit. For 2026 it's approaching $2T, if there is a decline in tax revenue it could go to 3+T easily. So they are doing everything possible to pump.
The Federal reserve doesn't have the balls to call it QE, now RMP, but same effect. Why anyone but the Fed is buying bonds longer than 1 year at these rates is beyond comprehension, but then again, so is TSLA stock price. The only possible future is yield curve control once people figure it out.
Western countries are in massive trouble. EU major contributors France and Germany are cooked. US is rapidly approaching a point where fixed expenses will exceed tax revenue, Australia and Canada have massive housing bubbles. Debt crisis is coming.
I sold nearly all of my stock earlier this year because AI sucks ass. It's obviously way over hyped and being pushed on everything rather than being pulled (I see you Microsoft). I bought some gold, need to add more now that short term rates have dropped. The only certainty is that I'm going to screw up and not make the right investments with my 401(k) and end up as a walmart greeter until I'm 85, if I'm lucky.
When this all happens, no idea, it will take a lot longer to play out than I want, but I'm expecting massive democrat gains in the next 2 elections. Any problem is going to be blamed on Trump, even though much of the above is decades in the making. Stopped by Sam's Club today on my way home from work and it was full of disgusting morbidly obese 3 foot tall latin women. It's fucking over very soon.
Of course all the problems will be blamed on Trump. It's always the classic problem we've been facing for decades. Even a total collapse won't change people's opinion. They'll say "if it weren't for Trump we never would have crashed". People are incapable, at a pure IQ level, to comprehend what's going on, which is why democracy is a terrible system.
The system can be fixed but I suspect they'll try all they can to avoid fixing it but also avoid letting it crash for another couple more decades or however long they can keep it going, which over the long term will cause far more harm than just fixing things or letting it crash.
Blame on Trump will be correct, though. He hasn't done a f***ing thing to fix any of the baseline economic issues, because a) hes a boomer and b) he continues to pick shitheads for staff
He thinks things are fine, and all his advisors are MMT acolytes who are feeding him garbage.
Medical costs are going to eat everyone alive in the next few years even if the economy doesn't crash harder than a jumbo jet from 41 thousand feet, and no one is willing to do anything about that, because D.C. is all bought and paid for by the medical lobby.
Correct. Trump's tariffs were right but he didn't cut costs (which is why Musk bailed on him). Costs need to be cut by 80% now but at the least they could have cut 10%. He expanded spending which was the worst thing he could have done.
The worst part about medical costs is people think the solution is tax more and spend more when it's literally tax less, spend less and let all the old people be unable to afford their care. They need to die.
Trump's tariffs were not right. If he'd just put tariffs on Chinese imports that would have been a good idea, but the way he did it was retarded.
Speaking of retarded, what's the news on the tariff "rebates"?
It should be blamed on Trump. He failed at making cuts and actively increased spending for his Christian agenda.
MAGA:"Did someone say more birth subsidies?"
Ok, I have a different answer. Let's say AI is a big change. It rarely goes in the direction the people in charge want. They want a closed group they control and Lord over everyone else.
Imagine 100 years ago. The Model T is just coming out and it's starting to sell well. Movies are starting to make it big. What does everyone base things on? That's right, what they're accustomed to. So, cars are for upper class and maybe upper middle class people. Movies are a spectacle and that's it. The masses don't need it. There is industry entirely aimed at the markets for this. People were paid to dig up the crap off the street.
If you told an industrialist from that era what society looked like, would he understand? If you understood that the change is happening, how would you help move it? How would you help those that need help? As a boy scout I was paid to dig crap out of barns and streets. It did not pay well since so little was needed.
Everything is aimed at industries that won't matter, or won't be used the way everyone thinks. Money is being used to prop up businesses and industries being hit, but no one knows what to do beyond that.
The guy who came up with the term Disruption said that Silicon Valley and tech central cities would become the next Detroit. Seattle has a lot of rich industrialists who want centralized power, but no real idea where anything is pointing. Meanwhile the area itself grows poorer and more insane.
Get this, Detroit still has steel mills. In fact they make more steel than they had before. The technology itself needs way less people. The city was designed for the jobs to be permanent.
Are we seeing Billions and even trillions going into a direction that may or may not be the future? Let me correct myself. Let's say AI really is the future. Is the money being spent right now for it to happen where AI will actually end up? We saw people aim VR in the worst direction. It's not that it is bad, it's that the very same people promoting AI right now did not understand what they were aiming for. They built a city for jobs that didn't exist.
The best answer is to figure out how AI and automation will actually be used. Elon Musk tends to be a big industrialist. I think the space data centers are just his excuse to get us into space. His aim is space, AI and all that are just a means to get there. I've had a lot of people tell me to do things in a specific way because it's just what they did beforehand with AI. Sort of like how the early movies were just Broadway plays with cameras. The early cars were just the wagon without a horse and then placed an engine. Model T didn't have foot pedals the way you know them. The original pinball table looks sort of like a boardgame. It was placed in an industry the way most would expect to work, but isn't what it will end up being.
Theme Parks are very expensive and teaching their limits. Conventions are cheaper and change faster. Pop up Experinces can be anywhere and don't need the cost and insanity of the convention. Make a YouTube show and other stuff and you have the modern entertainment. Does anyone who designs theme parks know this? Yeah, I know a few. None of them want it to hurry up because they're out of a job and their skill set is not as needed.
So, I think the money is aimed at people trying to get power or keep an industry going. No one in power knows where it is going, and those that so will try to stop it if they can. All this money is for keeping jobs and power.
Don’t matter. Don’t exist. Stop talking about TV shows as though they’re real.
The only thing that matters is the purposeful crash they’re going to cause in order to make all currency illegal and force people into a single global CBDC. Timing is irrelevant because no one will do anything about it. Won’t be long, though. All their pilot programs worldwide are complete.
"AI" in the consumer sense will be a major disappointment for 90% of people. 10% will get benefits from AI doing a lot of filler work. The vast majority of real value of these datacenters is furthering along the surveillance state and the massive amount of money that having perfect data packages (allowing for hyper-tailored ads or user experience changes that are perfectly consistent across the internet) for a third of the people in the world is insane and worth every penny for globohomo.
Yeah, it's all about th surveillance state but I still don't think current valuations are pricing the surveillance state properly here. Way too overvalued, which might not matter if they can get the government to pay for it using the money of the people being spied on. But this sort of lack of productivity has consequences that can't necessarily be hidden forever before there's a blowout.
It can be both, like how the dotcom bubble crash was followed by everything gradually moving online.
There is a rumor going around In the tech community IE Gamers Nexus that following the bubble pop, the data centers will be re purposed for cloud computing as personal computers will become prohibitively expensive.
Dotcom helped spur on the internet stage because even after so many tech companies lost everything, the underlying infrastructure was still there and ripe for exploitation.
Because the sane one focused on books at first, and used that to build out an infrastructure that could have been done prior to the internet, but that no one else(looking at you, Sears) was willing to do.
It is a bubble thats being propped by their own funds though. Its a deal between microsoft, nvidia, oracle, intel, amd, and openAI, among other smaller actors.
Trump did allocate funds that were already "spent" on paper but not spent literally.
Im not sure if it is going to cause a crash for these reasons, although it most certainly is a propped up bubble. In the long run, I believe this bubble is intended to push us towards building data centers, and it may wind up pushing us towards more reliance on nuclear energy and perhaps if were lucky they will be forced to use an efficient design, like the molten sand reactor.
If you dont already own a pc that can locally run AI, though, you might not be able to purchase one in the future.
I should also mention I spoke to some of the CEOs of SMR nuclear power plants and they are all israelites.
If true then it's almost a definite then.
AI is amazing at programming.
Re AI: Can you describe to me the business use case for it? how does a business monetise AI in a way they can profit from?
That's the issue with AI, it can't be monetized for long. It will become commonplace and free like a .rar extractor program.
I think AI companies have realized this and their only ploy is create an AI that demands so much hardware power to make that it can't be emulated for free but I have a feeling this bet is going to fail, which is why their second hedge seems to be to buyup all the hardware so people can't even run dedicated AI for free even if they wanted to. This may extend the timeline for how long they can charge for AI use but soon enough the information will be there for people to make and run their own free dedicated AI programs, which will be preferable because they won't be censored or tracked; however, the lack of control scares the government so they'll do all they can to make this not a reality.
The purpose of the current push for AI is to replace labour costs, either by outsourcing directly to AI or by using AI to prop up your previous outsourcing scam by making your Indians less incompetent. There are other use cases - machine vision, machine translation, etc - but that's small fry in comparison.