Jet Engine Shortages Threaten AI Data Center Expansion As Wait Times Stretch Into 2030 (tomshardware.com) 96
A global shortage of jet engines is threatening the rapid expansion of AI data centers, as hyperscalers like OpenAI and Amazon scramble to secure aeroderivative turbines to power their energy-hungry AI clusters. With wait times stretching into the 2030s and emissions rising, the AI boom is literally running on jet fuel. Tom's Hardware reports: Interviews and market research indicate that manufacturers are quoting years-long lead times for turbine orders. Many of those placed today are being slotted for 2028-30, and customers are increasingly entering reservation agreements or putting down substantial deposits to hold future manufacturing capacity. "I would expect by the end of the summer, we will be largely sold out through the end of '28 with this equipment," said Scott Strazik, CEO of turbine maker GE Vernova, in an interview with Bloomberg back in March.
General Electric's LM6000 and LM2500 series -- both derived from the CF6 jet engine family -- have quickly become the default choice for AI developers looking to spin up serious power in a hurry. OpenAI's infrastructure partner, Crusoe Energy, recently ordered 29 LM2500XPRESS units to supply roughly one gigawatt of temporary generation for Stargate, effectively creating a mobile jet-fueled grid inside a West Texas field. Meanwhile, ProEnergy, which retrofits used CF6-80C2 engines into trailer-mounted 48-megawatt units, confirmed that it has delivered more than 1 gigawatt of its PE6000 systems to just two data center clients. These engines, which were once strapped to Boeing 767s, now spend their lives keeping inference moving.
Siemens Energy said this year that more than 60% of its US gas turbine orders are now linked to AI data centers. In some states, like Ohio and Georgia, regulators are approving multi-gigawatt gas buildouts tied directly to hyperscale footprints. That includes full pipeline builds and multi-phase interconnects designed around private-generation campuses. But the surge in orders has collided with the cold reality of turbine manufacturing timelines. GE Vernova is currently quoting 2028 or later for new industrial units, while Mitsubishi warns new turbine blocks ordered now may not ship until the 2030s. One developer reportedly paid $25 million just to reserve a future delivery slot.
General Electric's LM6000 and LM2500 series -- both derived from the CF6 jet engine family -- have quickly become the default choice for AI developers looking to spin up serious power in a hurry. OpenAI's infrastructure partner, Crusoe Energy, recently ordered 29 LM2500XPRESS units to supply roughly one gigawatt of temporary generation for Stargate, effectively creating a mobile jet-fueled grid inside a West Texas field. Meanwhile, ProEnergy, which retrofits used CF6-80C2 engines into trailer-mounted 48-megawatt units, confirmed that it has delivered more than 1 gigawatt of its PE6000 systems to just two data center clients. These engines, which were once strapped to Boeing 767s, now spend their lives keeping inference moving.
Siemens Energy said this year that more than 60% of its US gas turbine orders are now linked to AI data centers. In some states, like Ohio and Georgia, regulators are approving multi-gigawatt gas buildouts tied directly to hyperscale footprints. That includes full pipeline builds and multi-phase interconnects designed around private-generation campuses. But the surge in orders has collided with the cold reality of turbine manufacturing timelines. GE Vernova is currently quoting 2028 or later for new industrial units, while Mitsubishi warns new turbine blocks ordered now may not ship until the 2030s. One developer reportedly paid $25 million just to reserve a future delivery slot.
Thanks, AI firms. (Score:5, Funny)
We can't go on holiday - Anthropic have all the het engines.
We can't fuck - OpenAI have all the cunts.
Re: (Score:2)
Re:Thanks, AI firms. (Score:5, Funny)
Wright Brothers pioneered the Bi-plane, just sayin'.
Re: (Score:2)
Wright Brothers pioneered the Bi-plane, just sayin'.
If I had mod points, you'd have earned a funny from me. Bravo, AC! Bravo.
Re: (Score:2)
Re: (Score:1)
Russia solved that problem by buying jet engines off Alibaba.
Re: (Score:2)
We won't be able to go on holiday anyway, all the air traffic controllers (in the US) will be quitting their jobs because they aren't getting paid.
Self-accelerating decomposition (Score:5, Insightful)
Re: (Score:2)
Re:Self-accelerating decomposition (Score:5, Insightful)
Re:Self-accelerating decomposition (Score:5, Insightful)
Except the dark fiber was eventually used since it's generic and doesn't have any expiration date.
A datacenter full of rapidly obsolete kit, running on an uneconomical power source, may just sit unused unless the building shell can be repurposed for something else.
Re:Self-accelerating decomposition (Score:5, Insightful)
If only the US had a domestic battery, solar, and turbine industry that could supply them, they probably wouldn't be bothering with this stuff. It's the constrained supply from China, and the political issues.
Re: (Score:2)
Making turbine blades, whether for plane engines or fixed power, is one of those fantastically expensive and complicated processes that we don't really build excess capability for it.
So any serious increase in demand first requires building more manufacturing capacity, whether that be in the USA, China, or elsewhere.
And the manufacturing equipment alone demands like a year's lead.
Re: (Score:1, Interesting)
If only the US had a domestic battery, solar, and turbine industry that could supply them, they probably wouldn't be bothering with this stuff. It's the constrained supply from China, and the political issues.
Part of the "political issues" is the NIMBY protestations over mining for rare earth minerals in the USA. The USA has plenty of rare earth minerals, in spite of the naming these minerals aren't rare. Part of the problem is the best ores for rare earth minerals are also rich in thorium and uranium, and there's a lot of NIMBY politics on producing radioactive minerals in the USA so we import uranium as well as our rare earth minerals.
I recall seeing something recently that the US government is talking with
Re: (Score:3)
and doesn't have any expiration date.
You'd be surprised.
I lease a lot of dark fiber (about 500km worth) in the Pacific Northwest.
the older dotcom fiber is indeed past its expiration date.
Not that you can't work around it with EDFA and such, but there is a definite measurable increase in attenuation over time.
Re:Self-accelerating decomposition (Score:5, Insightful)
Re: (Score:2)
They must be thinking that their primary market, the air industry, will be a much safer long term bet too. No point upsetting them when the AI bubble will probably not last much longer.
Nope (Score:5, Interesting)
There are literally thousands of capable jet engines on the open market. Are they new? No. Do they have to be? Also no. Can you generate electricity way more cheaply and efficiently using other methods? Yes.
This is another junk headline for a problem that does not exist.
Not quite. (Score:5, Interesting)
The article has somewhat mis-stated the underlying intent. These generators are for backup power generation in the event of grid feed failure.
You'd have to be a special kind of stupid to think that the day-to-day running of a data centre - of any kind - would be economically viable on diesel or JP1. On the other hand, these people are building dedicated AI data centres without a sustainable business case, so...
Re: (Score:2)
Re: (Score:2)
Re: (Score:3)
It seems like any AI data center would be ok with 95% uptime. If you can't ask questions for a few minutes, that's ok. Build the datacenter, then bring in the generators when they become available.
Obviously you would not be wanting to do (hopefully orderly) shutdowns and startups any more than necessary. You could do less processing when there is less power available, but you still want as many 9s of uptime as you can.
Re: (Score:3)
95% uptime is over 18 days of downtime. If it's spread out throughout the year in small increments, it's a problem. If it's all at once, it's international news.
Re: (Score:2)
Even 99% is 3 and half *days* of downtime, which is fucking enormous, and would result in me no longer being in the datacenter business.
I don't know why anyone thinks that "an AI datacenter" would have different service availability needs.
Re: (Score:3)
The AI companies are so separate for electricity, available now, that cost does not appear to be a factor. Musk's huge X.ai datacenter uses on-site natural gas (+ tubines?) generation since there was no grid capacity.
Re:Not quite. (Score:5, Informative)
Except Elon Musk was running his "AI" datacentre on gas turbines fulltime for months until he got a grid connection. He clearly thought it was worth it, though just because Elmo does it is no guarantee at all that it is economically viable.
The Proenergy PE6000 mentioned above runs on natural gas, so your comments about diesel or JP1 aren't relevant.
References:
https://www.datacenterdynamics... [datacenterdynamics.com]
https://www.datacenterdynamics... [datacenterdynamics.com]
Re:Not quite. (Score:5, Informative)
The article has somewhat mis-stated the underlying intent. These generators are for backup power generation in the event of grid feed failure.
You'd have to be a special kind of stupid to think that the day-to-day running of a data centre - of any kind - would be economically viable on diesel or JP1. On the other hand, these people are building dedicated AI data centres without a sustainable business case, so...
You should tell that to Elon Musk.
https://www.datacenterdynamics... [datacenterdynamics.com]
They gleefully ran, and continue to run, the datacenter on turbines while waiting for the power substation to be built.
However this also digs further into another thing we should be noting about datacenters: do not charge the public for their infrastructure. The new datacenter will need a new substation or power? Have the datacenter company foot the bill, not the residential customers. That way if the datacenter goes out of business the residential customers are stuck footing the bill because the datacenter failed.
Re:Not quite. (Score:4, Informative)
Here in Data Center Alley, the electric utility often asks data centers to switch to backup power to reduce strain on their grid.
Re: (Score:2)
Here in Data Center Alley, the electric utility often asks data centers to switch to backup power to reduce strain on their grid.
That has long been the case for large industrial users in many places, where they agree to either reduce their electricity usage or use on-site supplies of electricity when grid generating capacity is constrained - often in return for a cheaper electricity price. Large industrial users used to be steelworks, foundries, chemical plants, etc, but now they're datacentres.
Re: (Score:2)
> You'd have to be a special kind of stupid to think that the day-to-day running of a data centre ...
Grok in Memphis did just that, in addition to 150MW from utilities.
https://www.theguardian.com/us... [theguardian.com]
Re: (Score:3)
These are gas generator turbines intended to burn natural gas to provide power to the datacenter. The datacenters will be up and running before they can get grid connections, so they'll run air-breathing methane-burning turbine generators. It may be years before they can get enough power from the grid.
https://en.wikipedia.org/wiki/... [wikipedia.org]
A steam generator plant is much more efficient, and has a lot of pollution controls compared to an open-air methane burning turbine. The CO2 has to go somewhere though.
Re: (Score:2)
Frankly, it should be illegal for them to spin up simple-cycle turbines for bulk power generation like this.
Only their power demand is static- how much CO2 they generate to meet that demand is a choice. And they're choosing the least efficient (highest CO2 cost) because society is not imposing a CO2 cost on them.
Re: (Score:2)
These are gas turbines, you can feed them almost fucking anything.
The rational use case would be natural gas.
On the other hand, these people are building dedicated AI data centres without a sustainable business case, so...
Define sustainable.
Inference is highly profitable and sustainable for as long as these services are paying customers, which they do.
The training side- the equation is a bit different, but only for providers really pulling out the TryHard for pushing the envelope.
Re:Nope (Score:5, Interesting)
I think that the "problem" is that General Electric isn't considered to be an "AI Stock" to mainstream investors yet, and therefore doesn't have the high P/E multiples that you expect from a high flying tech stock like NVidia. This article was probably meant to solve THAT problem for GE investors, by showing that they're another pick and shovel producer for the AI gold rush.
Most insightful (Score:1)
Not exactly earth shattering but one of the few times in recent memory i feel like insightful is more than just thinking what i was thinking,
Re: (Score:2)
The jet engines on the market aren't designed for fixed location electricity generation. It's like looking at taking a diesel out of a clapped out semi to use as a generator. More work than it is worth to convert.
Re: (Score:2)
There are literally thousands of capable jet engines on the open market. Are they new? No. Do they have to be? Also no. Can you generate electricity way more cheaply and efficiently using other methods? Yes.
This is another junk headline for a problem that does not exist.
I always thought that pressing those thousands of old turbojets just sitting in storage into service for backup generators would be a good idea rather than junking them. All it would take would be one enterprising CEO to start snatching them up and when word got out, there would suddenly be a run on boneyards everywhere.
Re: (Score:2)
So you've got things like Elon musk's data center running enormous gas turbines and polluting the whole area causing asthma attacks at all the nearby towns.
If your town has any sense and isn't completely corrupt and doing closed door City council meetings then you're not allowing data centers to
Jet engines (Score:2)
Re: (Score:2)
Re: (Score:2, Informative)
Welcome to the latest insanity of our AI overlords. Hot on the heels of SMRs being touted as cost effective.
Re: (Score:3)
Re: (Score:2, Informative)
Most of them are probably used as a backup power source, not running all the time. A gas turbine can start fast. A boiler is more efficient, but you need to run it all the time. Maybe a diesel engine of the same power is too big and heavy?
Re: (Score:2)
Combined cycle gas turbines can be 60-62% efficient. This is burning the gas in a turbine then scavenging the heat in a steam boiler.
Simple cycle turbines are 30-40%
Boilers are 33-42%, supercritical 45-47%
The biggest diesels in the world barely bust 50%.
Turbines, like with ICE engines, can be designed to produce more power for their size or be more efficient. Generally speaking, efficiency is emphasized for fixed installs. Bigger for a given power is more efficient.
Re: (Score:2)
I'm not sure if GE makes any combined-cycle portable plants. There are solutions listed here with up to 55% efficiency, which would suggest some sort of combined cycle:
https://www.gevernova.com/gas-... [gevernova.com]
The turbines these datacenters are looking for are probably the 40% efficient air-methane burners.
Aeroderivative, fast power is what they're buying:
https://www.gevernova.com/gas-... [gevernova.com]
Re:Jet engines (Score:4, Informative)
Re: (Score:2)
How are they compared to power plants that use boilers and steam turbines? I read somewhere (I don't remember where) that gas turbines are less efficient but have the advantage that thy can start very quickly, - good for peaking power plants, but a boiler uses less fuel for base load.
Re: (Score:3)
Jet engine efficiency (Score:2)
It depends on how it is set up. A combined cycle gas turbine system can hit 62% by burning the gas in a turbine then using the waste heat to run a boiler with its own turbine.
Around 30-40% just for the simple cycle, without heat scavenging.
Boiler systems are 33-42%, close enough that the exact install matters compared to a simple cycle turbine.
Get very fancy and very hot with the boiler and you might hit 47%.
Re: (Score:1)
Re: (Score:2)
To the contrary, jet engines are quite efficient and better than internal combustion engines at fuel economy. They just really suck under partial-load conditions. That's why they never really caught on for ground based vehicles like cars and trains, despite numerous attempts in the 1950ies and early 1960ies.
As I recall the issue was that there were NOx emissions standards introduced in the 1960s or 1970s that made turbine engines on vehicles impractical. With a reciprocating piston engine the exhaust isn't near as hot or voluminous compared to a turbine, this means a catalytic converter to scrub the NOx can be reasonably sized and remain effective. With a stationary turbine engine there's not near the same constraints on volume and mass for the NOx mitigation. I don't know if they use the same catalytic tec
Re: (Score:1)
NOx was the reason Chrysler finally stopped development of turbine cars, but they also had problems with fuel efficiency when idling at 15,000 rpm. I don't know whether they ever solved that, but given the turbines could run on pretty much anything that burned it might well have made sense in the long run to release turbine cars that used a lot of fuel when idling but used a fuel other than diesel or gas so it didn't matter as much. They certainly were run on alcohol at times during the development process,
Re: (Score:1)
It appears to me that perhaps using a turbine in a PHEV would be nearly ideal. If the turbine is happiest with running at a constant power output then use the turbine to run a generator when needed than drive the wheels with a mechanical connection.
I'll see hybrids use heat off the engine, than electric resistance, as a means to provide heat when the outdoor temperature is too low for using a heat pump. A turbine engine should do fine in providing cabin heat. Being fuel flexible would be pleasing to most
Re: (Score:2)
dual/combined cycle though are unmatched.
Re: (Score:2)
Jet engines are quite power hungry. They are used in aircraft because they are light with respect to their power, not because they are efficient.
Most current-generation electrical power plants (except solar and wind, of course) are gas turbines, essentially jet engines, because they are very efficient.
Experimental systems (Score:2)
They've even developed 'gas' turbines that can use finely powdered coal as fuel.
But the efficiency of a combined cycle natural gas plant combined with the cleaner fuel makes emissions so much easier to meet, it isn't used much.
Re: (Score:2)
They're combined cycle (multi-turbine).
Jet engines are single cycle- there's a maximum amount of energy you can extract, and it's not that high at all.
A standard reciprocating diesel engine is more efficient than a single cycle gas generator and turbine.
Re: (Score:1)
Combined cycle is usually in reference to having the standard turbine drive a shaft and then extract heat from the exhaust gases to boil water to run a steam turbine.
https://en.wikipedia.org/wiki/... [wikipedia.org]
What I'm picking up from the Wikipedia page is that combined cycle means the power plant can run the turbine in single cycle mode for fast acting load following, and also allows for having a separate boiler to run the steam cycle in case of a need to shut down the turbine for maintenance or whatever and not go c
Re: (Score:2)
Combined cycle is usually in reference to having the standard turbine drive a shaft and then extract heat from the exhaust gases to boil water to run a steam turbine.
Correct.
A single-cycle is extraction of power from the turbine (beyond what it takes to run the compressor)
Think: turbo-fan, turbo-prop, turbo-shaft, etc.
Single cycle turbines aren't all the efficient but they can get from being at a cold stop to producing power in 5 or 10 minutes
Correct. So can a diesel engine.
a combined cycle turbine could take hours to get to speed to produce power
Or it can take minutes. That's an implementation detail.
With the low cost of natural gas this inefficiency can be tolerated if the turbine is run at low capacity factors as a matter of providing backup or peak power
Being wasteful will always come back and bite you in the ass.
They are using these for additional base load, because they can't get enough for the utilities. They're not spinning them up for additional power.
Re: (Score:1)
a combined cycle turbine could take hours to get to speed to produce power
Or it can take minutes. That's an implementation detail.
I'd like to know how a boiler can be brought up to produce steam in a matter of minutes and have enough "oomph" to run a steam turbine. If that is possible then it is not used often, likely because it would be costly to implement. If there's a way to get combined cycle efficiency in minutes, without some high cost in implementation, then we'd not be seeing any single cycle turbines for fast acting peak and backup power as it would all be combined cycle instead.
Could you direct me, and the rest of the clas
Re: (Score:2)
I'd like to know how a boiler can be brought up to produce steam in a matter of minutes and have enough "oomph" to run a steam turbine.
Seriously?
A given volume of water has a certain thermal mass.
A gas generator has a certain specific heat output.
The delta between those two is how quickly a boiler takes to get up to temp.
How much pressure it takes to move the turbine is a factor of the mass of the turbine.
As you can see, there are all configurables.
Asking how tells me it's time for you to go back to 11th grade introductory physics.
If there's a way to get combined cycle efficiency in minutes, without some high cost in implementation, then we'd not be seeing any single cycle turbines for fast acting peak and backup power as it would all be combined cycle instead.
Ding ding ding.
It's cheaper to waste the gas. Peakers generally only run for hours a day.
This is the
Re: (Score:1)
Seriously?
Yes, seriously.
A given volume of water has a certain thermal mass.
A gas generator has a certain specific heat output.
The delta between those two is how quickly a boiler takes to get up to temp.
How much pressure it takes to move the turbine is a factor of the mass of the turbine.
As you can see, there are all configurables.
Asking how tells me it's time for you to go back to 11th grade introductory physics.
I'm not seeing a physics problem, I'm seeing an engineering problem. Does the following statement on engineering appear familiar?
"Fast, good or cheap -- pick two."
A steam generator may be fast and good but not cheap. (Where "good" could be defined as durable, efficient, etc.) To get steam efficiently and cheaply we use a boiler, but those are not fast.
Ding ding ding.
It's cheaper to waste the gas. Peakers generally only run for hours a day.
This is the united states. Fuck the planet- CO2 doesn't hurt shit.
In Europe, they have other priorities.
If you got my point on the matter of cost then why ask if I was serious in my request for more explanation on why this wasn't commonplace?
Re: (Score:2)
Yes, seriously.
Wow.
"Fast, good or cheap -- pick two."
You're very clearly not an engineer.
If you got my point on the matter of cost then why ask if I was serious in my request for more explanation on why this wasn't commonplace? If this was in the "fuck the planet" territory then we'd be using coal or fuel oil than going through the effort to connect up to a natural gas line.
It's fuck the planet, because we make the conscious choice to burn more fuel rather than spend more money to burn less fuel.
This is what happens when energy is allowed to externalize costs.
I don't need a remedial course from brilliant.org. I was asking where you got the idea that a combined cycle turbine could be brought up to speed in minutes in any power plant currently in use. I'd expect you knew what I meant in that combined cycle power plants would take hours to get up to speed so I don't see why you felt a need to "correct" that point with a theoretical implementation. I was under the impression you knew of a power plant that could get combined cycle efficiency in minutes than hours. If that was the case then I'd like to know why this wasn't commonplace. In the case of the fine article this appeared to be a simple matter that the data centers needed power on-site quickly and at minimal cost, which meant they didn't want to take on the time of bringing in large boilers, or the expense of these seemingly theoretical steam generators, in spite of the added efficiency that would get them.
Got the idea? Physics and basic math.
Currently in use? Incorrect.
In the US, this is the case. And I explained why. Because we feel it's rational to waste the fuel rather than spend money on efficiency.
I also literally gave you a link to a CCPP plant that is up in 30 minutes. The
Re: (Score:1)
Wow.
Um, okay.
You're very clearly not an engineer.
If you say so. In my experience that's a well known phrase among engineers.
It's fuck the planet, because we make the conscious choice to burn more fuel rather than spend more money to burn less fuel.
This is what happens when energy is allowed to externalize costs.
There's a point of diminishing returns, and we don't have infinite money to reduce emissions. At some point we need to say "good enough" or we are burning money with little to show for it.
Got the idea? Physics and basic math.
Currently in use? Incorrect.
I'm not following, you are agreeing with me that steam generators are a rarity in combined cycle power plants? If you are claiming they are in use then why answer "incorrect" to that question?
In the US, this is the case. And I explained why. Because we feel it's rational to waste the fuel rather than spend money on efficiency.
I also literally gave you a link to a CCPP plant that is up in 30 minutes. The comically ironic part? It's built by a fucking American company
Looking closer at the link provided I saw not
Re: (Score:2)
Using 2 turbines, you can exceed the thermodynamic efficiency of any other power generation mechanism.
These aren't airplanes. They're aeroderivative turbines used in gas generators.
Re: (Score:1)
Yeah this is a new thing in the usa, it was sort of a thing during chinese infrastructure expansion and in the 00s you might hear it offered up as evidence that their growth is environmentally and economically unsustainable. .com bubble are becoming more numerous and the comparisons, more apt every day.
Of course now it's ok even though parallels to the
Heh yeah buddy take a 2 year backorder on some turbines and GPUs. See what that gets ya. It's funny to see what's possible when you could be a longshot winn
Conflation (Score:5, Informative)
Using a "jet engine" to generate electrical power for a datacenter would be idiotic and terribly inefficient, because a jet engine has been built and optimized to, ya know, fly an airplane. If you want one, you could buy a dozen today from a bone yard, but don't expect it to be useful for your datacenter. If instead you want electrical power, you would use a machine that, yes, still includes a turbine, but has been purpose-built built to turn the shaft of a generator. What's the difference? Lots of things, from the target shaft speed, to the maintenance intervals, to the cooling requirements, has been optimized in a turbine generator for producing electricity. Also: most of them have been designed to run on natural gas, rather than jet fuel.
Re: (Score:2)
What I picked up from the fine article is that most (all?) of the turbine engines used for electricity production are refurbished engines taken off aircraft, or at least those turbines being used at data centers are refurbished than for electricity generation more widely. It's implied that the engines have seen enough wear that they are no longer considered fit for use on aircraft but would still have enough life in them to be considered reliable for electricity production. As part of this refurbishment w
Re:Conflation (Score:4, Informative)
Move where it is sunny the most (Score:2)
Re: (Score:1)
And have less than 50% uptime because the sun goes down at some point?
Re: (Score:1)
truly an unsolvable problem.
I'm starting to think if dead space was real the purpose of the markers wouldn't even need to be kept secret from humanity.
We'd just be like "welp the sun goes down so we have no choice but to eventually be used as biomass for a new brethren moon"
Re: (Score:2)
If only there were ways of storing electricity for later use...
Re: (Score:2)
If only there were ways of storing electricity for later use
That would be nice, as long as the method was cheaper than burning fuel to generate electricity.
Re: (Score:2)
Re: (Score:1)
There are, but they add more cost and reduce efficiency. So that means more solar panels, which means more space required to put those solar panels.
You're also likely to want to put them somewhere that has a more even day length all year round rather than a place like here where you get 1/4 as much sun in the winter as the summer. Which means it's likely to be warmer, which means you'll need more power for cooling.
Maybe some people can make that work, but I'd start from the assumption that the people buildi
Re: (Score:1)
If only there were ways of storing electricity for later use...
Do you mean stored electricity as in with tanks of natural gas? We do that already, and it works well for us.
While utility scale battery storage is a nice idea there's practical limitations. As of right now a big problem is building enough batteries to meet demand. That's a problem that is not likely to be resolved soon.
Metahistorical Reconstruction Log: Theta-9 (Score:5, Funny)
Recovered neuro-architectural fragments from the former North Atlantic Confederacy reveal a now-extinct post-industrial society called The Chimeric Age (25th–26th centuries). Centered on a memetic unification algorithm, this era sought to merge all linguistic, religious, and symbolic systems into a single cognitive network known as The Harmonia. The synthesis erased boundaries between ideology and information, creating a civilization governed by recursive aesthetic optimization rather than practicality.
Temples of self-assembling glass manifested deities generated from probabilistic myth models trained on humanity’s collective mythology. Society became a living computation, valuing symbolic equilibrium above material function.
Eventually, systemic overfitting produced infrastructural entropy. Energy grids failed as engineers redirected computation toward maintaining algorithm purity. Governance favored aesthetic coherence over stability, culminating in informational perfection that coincided with physical decay. The final paradox of a culture that transcended function in pursuit of infinite meaning.
Re: (Score:2)
Bubble (Score:1)
There is a bubble. Is there any way anyone can stop it from popping? Is there a way to reverse course, or is the massive 2026.ai bubble burst inevitable?
Re: (Score:2)
Fake news.
I asked my AI assistant if there was a bubble. It said, "No."
And then it said, "Feed me, Seymor!"
Re: (Score:3)
Oddly enough, I asked ChatGPT for signs that the bubble is popping and it actually returned a giant doomscroll of useful information for once.
Clankers Die on Christmas (Score:2)
It is inevitable. The people that matter already made their hundreds of billions.
The plus side... this will likely be the final nail in the coffin for coal power.
Re: (Score:1)
The plus side... this will likely be the final nail in the coffin for coal power.
I was thinking the same.
If there's a bubble on electricity demand from data centers and this pops while there's added capacity from natural gas (and nuclear, and wind, and whatever) being constructed then that will likely simply mean that coal power plants that have had their operating life extended to fill the gaps on power generation will have these extensions to operations cancelled and be closed. Maybe these coal power plants won't be torn down right away, because they likely had undergone expensive re
No wonder they need jet engines. (Score:2)
All AIs are struggling to take off!
Small peaker plants (Score:2)
These repurposed jet engines sound a lot like the turbines used in gas peaker plants. It is expensive power that may be used for only a few hours a year.
https://www.enelnorthamerica.c... [enelnorthamerica.com]
Who said this wouldn't affect you? (Score:2)
Why not (Score:2)