Der CEO von IBM sagt, dass sich die Ausgaben für Billionen in KI-Rechenzentren bei den heutigen Infrastrukturkosten auf keinen Fall auszahlen werden
https://www.businessinsider.com/ibm-ceo-big-tech-ai-capex-data-center-spending-2025-12
Der CEO von IBM sagt, dass sich die Ausgaben für Billionen in KI-Rechenzentren bei den heutigen Infrastrukturkosten auf keinen Fall auszahlen werden
https://www.businessinsider.com/ibm-ceo-big-tech-ai-capex-data-center-spending-2025-12
25 Kommentare
“Stop talking sense man”….. bots probably
*AI stocks go up*
Also IBM: „Please come use our cloud environment, Pretty please?“ Also also, Watson!
At basically any cost. The market for this shit just isn’t there. Text, image, and video generation is cool tech; it’s a neat product. But I, along with 95%+ of the rest of the population, don’t have a use case for generated text or pictures or video, so how is this ever going to justify the investment? I’m never going to pay anyone money for a product that generates fake text, pictures, and videos because I have no use for them and I can’t even imagine a future where I would.
AI is worth maybe 0.1% of the claims made by the scam cultists
old man talk.
It means IBM cannot do it at all, not that others cannot.
And yet somehow will do the same thing tomorrow.
Large Language Mistake…
Oh come on how could making crap you tube video’s with AI not be a worthwhile endeavor when it takes all the power from any number of generating stations and untold amount of computer infrastructure so some dweeb can not use his brain.
Infrastructure costs will simply skyrocket cause these fools will pay any price to get rid of wage earners and crown themselves best boy.
It’s almost like renewable energy would have been a winning move. Oh well
Unless there is a serious breakthrough in the amount of costs involved it won’t make a profit. Not in terms of mass adoption?
Data centres and GPUs are not cheap to run.
Just continue to offshore the entire company to India and I’m sure you’ll get there 👌
also you’re going to have to come up with a backup plan for the millions of unemployed unhoused and unfed people rise up and start burning data centers.
I mean I’m sure the people that own most social media will gladly drive us into a civil war to try and bring those numbers down.
But it’s amazing to watch them speed up towards this cliff as if there is a plan for what to do when all these jobs are gone and none of the benefits trickle down.
I wonder what the total value of all labor is. If creating these data centers, chips, and all this infrastructure ends up costing more than it would to employ people for the same amount of time, then they’re really pushing the return on investment much farther out than I’m willing to bet they’ll tolerate. If more data centers and more energy yields stronger LLMs, which it doesn’t actually – you get diminishing returns, then their bets aren’t going to pay off at all.
However, the hope is that end to end AGI does come about as a result of all of this, in which case the bet does pay off, but only if the total labor value is less than the cost of its replacement. I don’t know what that number is but total labor value has to be less than 8 trillion dollars, I guess – what a goofy number btw. Yet, Krishna says that has between 0-1% chance of happening.
Bubble gonna bust, the cokeheads are gonna jump from buildings soon.
A SLAP – Solution looking for a problem
Too many people (CEO mainly) are trying desperately to shove the square peg called AI into round holes.
How is it that everyone knows it’s destined to fail and yet they’re still putting in more and more money into AI?
Feels like he’s right on the *today’s costs* part and wrong on the conclusion. Infra economics always look insane at the peak of hype, then unit costs collapse or half the players die. Also convenient take from a company that mostly missed the hyperscaler wave.
Why not get AI to build itself?
Well the people spending billions think this is the last invention that will lead to infinite profits….nothing else matters!
Don’t you dare affect that bubble yet. The rich need to make their money first.
The balls are already too deep and there is not enough room to wiggle
IBM is so far behind in everything it’s almost hilarious they’re still around.
Wasn’t it a chairman of IBM who once said the world probably only needed 5 computers
https://www.lexology.com/library/detail.aspx?g=164a442a-1b90-49e3-895d-4c54bb49ecce