Canals were at least solving a problem that actually existed and needed solving at the time when they were started. AI data centres are being built in anticipation of future demand, for use cases that haven’t been developed yet.
That’s not accurate from my understanding. I quit my SDE job in December, but my former coworkers say they use AI pretty much all day and find it useful. Ofc, the company’s systems were an indecipherable mess, mostly because of rushed choices those same people made everyday, but neither here nor there. This is why tech companies are cutting jobs.
Whether any of this is sustainable remains to be seen, but there are current use cases and real demand for “AI” data centers.
Current tech layoffs are mostly a result of over hiring during the pandemic. Blaming AI is just the sales pitch to investors try to prove that the AI spending was worth it.
Maybe. There were rounds of layoffs a couple years ago where companies used COVID over employment as the rationale. I think four years removed that rationale is a bit of a stretch.
I think it’s also possible that the looming recession is a cause, but they can’t say that out loud for fear of offending the Crybaby in Chief.
And if AI is working properly these “innovative” tech companies shouldn’t really need to fire anyone, right? All their bright employees should just be able to use AI to generate value for their respective companies.
Idk the answer, but I can say that I personally know high-performing engineers who swear by it. I mostly found it useful as a rubber duck, or to quickly tear down some code in a language with which I’m not acquainted.
I actively try to avoid AI for moral reasons, but I have found it useful for certain tasks. It’s not a panacea, but it’s also not useless.
A few years ago with ARM arriving at the data centers I envisioned there would be a day that density would go up and new data centers would be less in demand. I’m either too early or wrong.
I’m thinking you might have been wrong. Even without AI, I don’t foresee demand for compute going down. Even if everything went over to ARM, I think that would have just slowed the rate of new builds.
Working for a fairly large open source project and we began testing arm hardware for in house use and it was ridiculous how much compute we could fit in the same U space.
We were doing nothing with GPU though, so CPU and memory speed was our bottleneck.
We were definitely more compute dense with our arm cluster but we couldn’t get as large. Yet. Maybe it’s changed now
In economics, the Jevons paradox, or Jevons effect, is said to occur when technological improvements that increase the efficiency of a resource’s use lead to a rise, rather than a fall, in total consumption of that resource.
Unfortunately using less of a good thing isn’t how we do things.
Also known as induced demand. Most of a thing drives more demand for that thing.
ARM for big compute isn’t much more efficient than x86.
We will never need less processing though even if AI goes away.
At least canals are still useful to have around for recreational watersports.
The Internet has ruined me because I can’t read the word “watersports” without thinking of the fetish anymore.
Ruined?
Gooned*
from the Olde English “guined” —an easy mix-up
Right, that’s what the canals are useful for, what did you think they meant?
Well I am sure some canals are next to dogging sites
And transportation
Over 3,000 data centers are planned in the U.S., and I am not imagining they’re being built so everyone can generate slop to their heart’s content. Recent events like OpenAI becoming an official member of the military industrial complex and the F.B.I.’s deal with Flock make me think they’re largely being built as infrastructure for the digital cage.
Digital cage. I’m gonna make art using this prompt. Thank you.
The canal in my uni town was immediately replaced by the railways pretty much the instant it was completed. They’re pretty cool though - you can walk along it, go on a narrowboat, canoe, fish out shopping trolleys, hide bodies, catch dysentery, or get attacked by Canada Geese.
Railroads replaced canals in the Midwest because the canals weren’t large enough to handle a lot of cargo, unlike the Suez or Panama canals. The canals here are maybe 20ft wide, and locks were maybe 50ft long. Trying to modify them to handle more cargo would have cost tremendous amounts of money and also disable the canals while they were being worked on.
Trains were simply able to carry significantly more cargo to their destination in less time
The US profits a lot from having so many waterways. When they are usable, they literally are the cheapest and most efficient mode of transportation.
Differencr that railway is still the absolute best and most efficient way
Canals are more vastly more fuel efficient per kg of cargo. But rail is much faster.
Pretty terrible metaphor as realistically even without AI we will never need less compute. Compute is awesome, it’s a good thing to have more. The problem is the environmental damage not anything else.
That makes the metaphor imperfect, but the data centers are being paid with circular money that isn’t there, based on the assumption that AI will make obscene amounts of money to actually pay their bills when currently they’re losing lots of investor money to subsidize the price, and they’re already talking about transitioning to a usage based system, which might push some enterprises to use a FOSS alternative from China.
We’re in a bubble, when it pops it’s likely many data center projects will be abandoned before they’re done.
isn’t that a good thing though? let billionaires lose money then just don’t bail them out. Either way the datacenters will remain useful and might actually benefit real work. This is btw how we got a lot of open technologies where corporate overinvests and then community takes over.
Considering the MIC is in bed with AI corporations I worry they will get bailed out to some extent, and that the gains will remain completely private, with taxpayers being expected to foot the bill (or alternatively, dealing with the inflation while their wages remain stagnant). I hope you’re right though, I really want there to be a positive in the end, I’m just skeptical on this front.
The silver lining to me is that there’s encouraging developments in FOSS AI, developed mainly in China. Maybe after the crash we can get somewhat cheaper GPUs to run them locally.
I think it’ll be really hard to bail AI out on public funds with current reputation if bubble pops.
However I think real risk is not full bail but constant leeching of public funds to private enterprise. This is not new to AI and this datacenter issue is exactly that - have people bear the environmental costs for private gains of datacenters. The other argument that US “doesn’t need datacenters” is just bad - that’s not the issue, everyone will need more datacenters forever basically.
encouraging developments in FOSS AI, developed mainly in China.
I wouldn’t really trust China’s open source commitments though. FOSS is non-existent in chinese culture and it’s directly a competitive response to american corporate models. That being said, a good deed done for bad reasons still can be good!
Honestly if AI would gracefully decline now and we just had open source models we’d still have work for the next 10 years implementing these tools.
Historically speaking it’s most often your average citizen that pays for it in the end.
Ideally, we wouldn’t build infinite compute, just as much as we are actually using, and using it efficiently allows you to build less. We would still need datacenters even without LLMs, but they wouldn’t need to be so gargantuan, because even the worst, inefficient, nodejs-based, intern-written server you could ever encounter, would be heaps more efficient (or at least less demanding) than any LLM. This is true even from an economical point of view, or any practical point of view, not just environmental.
To quote Tannenbaum: “You know you have the right computer when you are always using 99% of it. If you are using 100%, you are being limited by the machine. If you are using 98% you have bought more than you need”. If datacenters were always running at 40%, we would build bigger ones.
It’d be like if the railroad went through the canal
The choice to compare data centers to canals rather than to railroads seems rather arbitrary.
It’s not arbitrary at all. In fact it’s the entire point.
Just think, after the systems-manufactured destitute addicts clean out the rare metals from those industrial pustules, the whole country’s houseless will have homes, right?
Right?
Is this post trying to insinuate that car-roads are more worthwhile than railroads? Or canals for that matter? Because they aren’t. You backed the wrong boom. Sticking with rail would’ve been better in the long run.
I think you did not get the analogy. The point was that there was a canal bubble in the US which was mostly a huge waste of effort because the technology was outdated before a network could be completed to turn useless parts into a useful network. This was neither the case for rail or road networks. Both technologies experienced bubbles and overconstruction but they yielded large networks for generations to come.