Per the article, _somebody_ is overvalued. On the one hand, at some point they will stop needing new hardware, and can drastically reduce costs; in this scenario, NVIDIA will hit a wall because they no longer need to buy hardware. On the other hand, NVIDIA can keep selling; in that scenario, Anthroic &c is doomed to continue purchasing hardware, so their costs can't come down.
Re:I'm Not Convinced the AI Bubble Exists
by otherme123 on Monday November 09, 2026 @02:15AM
(#50422133)
> Amortising data Center cost will take a long time. Within that time, if someone figures out how to make ai run on cheaper hardware, these data center cost will Become harder to recover.
Or datacenters are the old mainframes, and when they stop building them and squeezing the market, GPU prices go back to what they were before the frenzy or even lower, so anyone really interested can buy a decent GPU for a fair amount of money, and run models locally for peanuts. Meanwhile, dataframe owners go bankrupt.
Unless nobody is working on miniaturisation, then this reasoning might hold true.
As seen throughout history, any sufficiently absurd tech starts largescale (refrigerators, computer, radios) and eventually become portable.
Amortising data Center cost will take a long time. Within that time, if someone figures out how to make ai run on cheaper hardware, these data center cost will Become harder to recover.
> AI companies can also cut costs by no longer allowing people to use their products for free and temporarily suspending expensive AI research
That will just kill the industry.