• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Congress Should Start Planning for a Potential AI Crash Now, New Report Says

P

philb2

2[H]4U
Joined
May 26, 2021
Messages
3,411
https://law.vanderbilt.edu/congress-should-start-planning-for-a-potential-ai-crash/

Trillions of dollars of capital investment in AI infrastructure are propping up the U.S. economy, yet revenues from AI are not increasing at a rate commensurate to this investment. This overreliance on AI investment, coupled with the opaque and complex financial engineering involved, means that the cost of a market correction could look like 2008 — an economy-wide crash with unpredictable and systemic consequences. If such a crash comes to pass, Congress would likely consider legislation to reform the market. Instead of scrambling to identify the right policies during an economic crash, Congress should start thinking seriously about and planning for a potential AI crash now, according to a new report from Asad Ramzanali, VPA Director of AI and Technology Policy.


Ramzanali offers a set of proposals for post-AI crash reforms. These include:
First, Congress should curtail the financial engineering—circular equity investments, opaque debt, and distortive government subsidies—that may be the proximate cause of the crash, and the government should prosecute any related frauds and illegal activities.


Congress should reform AI markets by establishing a Glass-Steagall for AI, utility-style regulations for digital utilities, a new regulatory agency, and a ban on surveillance-based business models. Following the Great Depression, Congress passed Glass-Steagall, which was repealed in the late 1990s, setting the stage for the 2008 financial crisis. Today, fervent demand for AI software is driving investment in AI infrastructure (chips, data centers), but that investment is distorted by the fact that the same few companies often own or invest in both sides of the coin, driving overinvestment in infrastructure. C

I downloaded the full 48 page report. It's worth at least a skim. Scary stuff. I worry about a stock market "correction" that sandbags my 401K) I'm old enough to remember the 2008 recession and the impacts on the stock market.
 
And, as has been reported, many of these AI data centers are not being built due to lack of infrastructure equipment. This has "crash" written all over it. The potential market correction could be catastrophic.
 
madpistol said:
And, as has been reported, many of these AI data centers are not being built due to lack of infrastructure equipment. This has "crash" written all over it. The potential market correction could be catastrophic.
Click to expand...
The lack is probably slowing down the rollout which is probably a good thing. It means spending is limited.
 
madpistol said:
And, as has been reported, many of these AI data centers are not being built due to lack of infrastructure equipment. This has "crash" written all over it. The potential market correction could be catastrophic.
Click to expand...
Are they not being built, reason is due to lack of equipment? Or are they being built, but not due to lack of equipment?
 
Nobu said:
Are they not being built, reason is due to lack of equipment? Or are they being built, but not due to lack of equipment?
Click to expand...
I'm guessing it's that everyone wants to build datacenters so just like how computer hardware supply is harder to come by these days, so is everything else needed to build a datacenter, so building plans get stalled.
 
I'm worried less about the companies that will fail and more about where all the funding came from that they're raising even VC infusions of cash came from somewhere, so which house of cards is going to collapse as a result of this.
 
Noble Aquarius said:
It will be around 2028, Doing Astrology over 4 years, it will happen in 2028, The signs and patterns are there. as Astrology that propose that information about human affairs and terrestrial events may be discerned by studying the apparent positions of celestial objects
Click to expand...
Can confirm, I asked ChatGPT to read some tea leaves.
 
SecretStash said:
I'm guessing it's that everyone wants to build datacenters so just like how computer hardware supply is harder to come by these days, so is everything else needed to build a datacenter, so building plans get stalled.
Click to expand...
I was just a bit confused about the wording, not trying to imply anything.
 
The city I work for killed the the potential datacenter proposed in their area. General consensus of the people feel the data centers are bad. Drains on local water and electricity resources scares the public.
 
AI will never fail because it gives them unlimited spying powers in real time for everyone.
 
Let's ask ChatGPT when the crash will occur.


QUESTION: Based on current market and socioeconomic patterns, when do you predict the AI crash will occur?
Click to expand...

If by "AI crash" you mean a market crash in AI-related stocks and investment rather than AI technology disappearing, my estimate is:
The highest-risk window is roughly 2027–2029, not 2026.
Here's why.

What makes bubbles burst?​

Historically, technology bubbles don't burst because the technology is fake. They burst because:
  1. Capital spending outruns profits
  2. Revenue growth slows
  3. Investors realize future earnings won't justify current valuations
  4. Financing becomes more expensive
The current AI cycle shows several of those characteristics. Many analysts are increasingly focused on whether AI infrastructure spending is generating enough returns to justify hundreds of billions of dollars in annual investment. There is growing discussion around "AI capex fatigue" and the gap between spending and monetization.

Why I don't think the crash is imminent​

Unlike the dot-com era:
  • AI is already generating real revenue.
  • Enterprises are actively deploying AI.
  • The largest buyers are cash-rich hyperscalers rather than speculative startups.
  • AI is producing measurable productivity gains in some sectors.
That's why many researchers describe today's situation as "real technology with bubble-like pockets" rather than a pure speculative mania.

What could trigger the correction?​

The biggest trigger would likely be one of these:
TriggerProbability
Hyperscalers reduce AI datacenter spendingHigh
AI revenues grow slower than expectedHigh
Power and infrastructure constraints delay projectsMedium-High
Recession reduces enterprise AI budgetsMedium
Major AI regulationMedium
AI fails to deliver expected productivity gainsMedium
The most important metric isn't GPU sales or model quality.
It's ROI.
Eventually boards will ask:
"We spent $50 million on AI. Did we actually save or make $50 million?"
When enough companies answer "not really," investment sentiment changes quickly.

My probability estimate​

If I had to assign odds:
  • 25% chance: No major AI crash; valuations compress gradually.
  • 50% chance: 30–50% decline in many AI-related stocks between 2027 and 2029.
  • 20% chance: Dot-com-style collapse in AI infrastructure and speculative AI companies.
  • 5% chance: Crash begins before the end of 2026. Current spending commitments remain extremely large and many investors still see demand exceeding supply.

My base-case scenario​

The most likely outcome is not a single "AI crash."
It's a rotation:
  • Infrastructure winners become overpriced.
  • Datacenter builders and GPU suppliers cool off.
  • Many AI startups disappear.
  • A smaller number of profitable AI application companies emerge as the long-term winners.
That's very similar to what happened after the internet boom: the internet changed the world, but most of the early market darlings did not become the dominant companies of the next decade.
In other words, I expect an AI investment correction before I expect an AI technology collapse. Those are very different events.
Click to expand...
 
Nobu said:
Did they have a clear and robust plan for water and power services, and outline exactly who was paying for it and how much? Or did they focus on tax revenue and job growth?

If the latter, I would also be skeptical.
Click to expand...
What tax revenue though? Data centers that drive AI don't have huge tax revenue, or job growth. It's a very stationary number.
 
kac77 said:
What tax revenue though? Data centers that drive AI don't have huge tax revenue, or job growth. It's a very stationary number.
Click to expand...
If there is no tax revenue and no job growth, then there is no benefit to the community -- it increases their burden, takes land, uses resources, generates noise/pollution, and there is no tax revenue or job growth?

It's their job to convince us otherwise. If they can't, then of course it will be disapproved.
 
DogsofJune said:
The city I work for killed the the potential datacenter proposed in their area. General consensus of the people feel the data centers are bad. Drains on local water and electricity resources scares the public.
Click to expand...
Tell that to our Premier up in Alberta, she is all gun ho' about DC's being built in our province, ignoring, as always, what the experts tell her...
 
And chip stocks to the moon again, fucking Seagate at $1k! I don't have a single HDD in my lowly gaming PC anymore...can't wait for the epic meltdown to come soon enough. And ofc SNDK from 28 to 2200 in 1 month, how's that for Nvidia gain haters lol?
 
stashinsome said:
Part of me wants this crash to happen only because I want parts to get cheaper. But on the other hand, i dont want my wife to lose her job which is in tech. wish us luck
Click to expand...
If you're in tech you want AI to crash, this AI 'boom' has hurt a lot more tech workers than helped...the only ones who have profited from this nonsense are the snake oil salesmen at the top and the few folks in hw whose stocks have 20x overnight. Software which employed far more workers is down big and most of the real estate in bay area has followed apart from the swankiest areas like Palo Alto where all the chiefs live..
 
Last edited:
undertaker2k8 said:
If you're in tech you want AI to crash, this AI 'boom' has hurt a lot more tech workers than helped...the only ones who have profited from this nonsense are the snake oil salesmen at the top and the few folks in hw whose stocks have 20x overnight. Software which employed far more workers is down big and most of the real estate in bay area has followed apart from the swankiest areas like Palo Alto where all the chiefs live..
Click to expand...
well shit, i guess we can only hope then
 
Let’s just rebrand!

Data Center -> AI factory!

Sounds like a factory where lots of people work and stuff is made!! Everyone wants that!!
 
DogsofJune said:
The city I work for killed the the potential datacenter proposed in their area. General consensus of the people feel the data centers are bad. Drains on local water and electricity resources scares the public.
Click to expand...

The general public doesn't understand how datacenters work, or what the demands of one actually are. For example, I see so many people hyperventilating about the water thing, seemingly unaware that closed loop cooling is a thing. This whole opposition campaign has shades of previous opposition to nuclear energy. Then, as now, the only question these people don't ask is who is funding these campaigns because they will benefit by the blocking of said datacenters, just like the oil industry was funding a lot of the anti-nuclear sentiment decades ago. Critical thinking is absent in a lot of these discussions, but hey, they did their "research", which generally means they watched a video sent to them by their uncle on Facebook for a few minutes and now they're an expert.

To be fair to the other side, the tech companies need to do a better job of selling this to the public. Spending the first few months telling everyone it would replace their jobs, which they figured would drive investment from business owners and probably did, turned out to be a really stupid idea that a lot of them are now walking back. BYO power plants would also go a long way to assuaging public concerns.
 
HockeyJon said:
The general public doesn't understand how datacenters work, or what the demands of one actually are. For example, I see so many people hyperventilating about the water thing, seemingly unaware that closed loop cooling is a thing. This whole opposition campaign has shades of previous opposition to nuclear energy. Then, as now, the only question these people don't ask is who is funding these campaigns because they will benefit by the blocking of said datacenters, just like the oil industry was funding a lot of the anti-nuclear sentiment decades ago. Critical thinking is absent in a lot of these discussions, but hey, they did their "research", which generally means they watched a video sent to them by their uncle on Facebook for a few minutes and now they're an expert.
Click to expand...
Even with closed loop systems the water usage is high, tie in with power requirements, noise issues, many places do not need these. In Alberta we have some of the cheapest nature gas in the world, and that is the way they are trying to sell it here "Come build DC's, but you have to build your own gas power plants to power them", which sounds great, but with that deal, all other environmental survey's are waived! So they can destroy the local area and it is fine....

Then you get people who do not understand what DC's are and just think "Great, lets build them so we can keep our data in Canada"..........No, does not work that way, all the DC's proposed here are to be owned by U.S companies...and are Hyperscaler DC's....
 
I'm still waiting for an explanation for why they are so great for the community. I haven't heard a convincing argument yet, only "the fears you have are unfounded." Okay, great, but what good does the datacenter provide me? It doesn't pay me, reduce my burden, give me a job.... It just increases land value and therefore taxes I owe.
 
MrGuvernment said:
Tell that to our Premier up in Alberta, she is all gun ho' about DC's being built in our province, ignoring, as always, what the experts tell her...
Click to expand...

Which experts? Some of the dumbest ideas I read about come from self-described "experts". I've lost count of the number of news articles I've read that has "experts say" in the headline, only to end up reading something unbelievably inane.

Canada's full of empty space, cold weather, the most freshwater in the world, and relatively cheap electricity by global standards with world leading reserves in uranium and natural gas. Alberta, in particular, is an energy powerhouse. I have serious questions for any "expert" who says this isn't an ideal country for a datacenter buildout, but then I'm reminded that I live in Canada, a country that is increasingly terrified of ambition and would rather continue to watch the rest of the world advance into the future while we get comfortable selling houses and condos to one another.
 
MrGuvernment said:
Even with closed loop systems the water usage is high, tie in with power requirements, noise issues, many places do not need these. In Alberta we have some of the cheapest nature gas in the world, and that is the way they are trying to sell it here "Come build DC's, but you have to build your own gas power plants to power them", which sounds great, but with that deal, all other environmental survey's are waived! So they can destroy the local area and it is fine....

Then you get people who do not understand what DC's are and just think "Great, lets build them so we can keep our data in Canada"..........No, does not work that way, all the DC's proposed here are to be owned by U.S companies...and are Hyperscaler DC's....
Click to expand...

So build them in rural areas and stipulate that the AI companies have to fund power generation directly. Done. They're willing to do this. Google's buying their own SMRs and Microsoft bought in to Three Mile Island specifically to get access to more energy for their datacenters. These aren't difficult challenges to overcome.

Noise issues? Ok, so use Alberta's abundant empty space and build it there. You don't have to build one in a densely populated residential area in Calgary or Edmonton.

This isn’t just about data sovereignty, it’s about enabling our tech sector to embrace the future of technology so we can be global leaders like we were in the past with companies like Nortel and BlackBerry, but again, it’s Canada, so I suspect we’ll cancel everything and just build condos instead.
 
HockeyJon said:
So build them in rural areas and stipulate that the AI companies have to fund power generation directly. Done. They're willing to do this. Google's buying their own SMRs and Microsoft bought in to Three Mile Island specifically to get access to more energy for their datacenters. These aren't difficult challenges to overcome.

Noise issues? Ok, so use Alberta's abundant empty space and build it there. You don't have to build one in a densely populated residential area in Calgary or Edmonton.

This isn’t just about data sovereignty, it’s about enabling our tech sector to embrace the future of technology so we can be global leaders like we were in the past with companies like Nortel and BlackBerry, but again, it’s Canada, so I suspect we’ll cancel everything and just build condos instead.
Click to expand...

That's the kicker. If they were willing to take on all the cost associated with these data centres it wouldn't be as big a problem. But, they want to build close to towns so they have easier access to power and water but then the noise pisses everyone off. If they were all willing to go build in the middle of nowhere and pay for the utilities to get to them the whole script would be flipped and much of the opposition would vanish. Blaming the opposition on "foreign actors" (the most recent push) is forgetting that they're pissing off actual people in actual towns where they got them built before the objections started escalating. People filming in their back yard about the noise, or people having actual power bills jump up is not some great foreign conspiracy and the more they push that narrative the more they'll distance themselves from the greater population. All they're doing right now is fueling the anti-AI movement with some small exceptions. Google and Microsoft have been pushing efforts to minimize power and water impacts and started pushing that narrative. Open AI, on the other hand, is rambling on about foreign conspiracies.
 
1781542537813.png
 
I mean around me most of the places here haven't moved to underground power/data yet so it does look like that cartoon. Except usually it is just trees hanging against the wires.
 
Nobu said:
My owned asset doesn't pay the taxes, I do, and increased land value doesn't increase my income.
Click to expand...
Yeah, owned asset value doesn't do anything to help me unless I'm borrowing against it or selling it. If I'm borrowing against my house I probably have more financial problems than the earned value on my home is going to help with.
 
You must log in or register to reply here.
Back
Top