4 ms·
What's driving the need for datacenters All over the world? Aren't there only a handful of companies need compute and can build datacenters? This isn't aws bui
by tayo42 4mo ago
What's driving the need for datacenters All over the world?
Aren't there only a handful of companies need compute and can build datacenters? This isn't aws building more regions is it?
Also recently I was surprised and not surprised to find people are making anti ai like their identity now. Like there's a reddit community dedicated to this
- ChiperSoft 4mo agoInvestors, thats what is driving it. AI is the only thing venture capital has cared about since 2022, so the people who earn fortunes from sucking on the VC teat are following the money. It doesn't matter if the data centers are worthless in five years, they will have extracted wealth from building them.
- SoftTalker 4mo agoHow do you extract wealth from building a highly capital-intensive physical plant? Construction crews and building material suppliers do not work on a promise of future payments.
- coldpie 4mo agoYou get VC to give your company $X00 M to build out AI capabilities, take a $X M salary for yourself, and retire. Doesn't matter what happens past that.
- ChiperSoft 3mo agoThe same way subcontractors make money, they take a cut of the building cost. VC invests a billion on a five year contract, you take a million as your salary for those five years and use the rest to pay for construction and initial operating costs. If the data center isn't profitable, oh well, you got paid 5 million all the same. If it is profitable, well thats just a bonus.
- jeroenhd 4mo agoInvestors chasing the ideal of not having to employ any people while still making money somehow. And, of course, being the company to provide all the AI; in a gold rush, don't dig for gold, start selling buckets.
- tayo42 4mo agoI get that, I mean who are the investors? Random colo companies? Are The big cloud providers are bringing up new small regions in all of these random towns? Only so many companies can even get the hardware to fill it. And there are only a handful llm providers. There's only so many companies involved as far as I can tell,but these datacenter stories seem everywhere?
- Spooky23 4mo agoThere’s a whole layer of companies providing these data centers. CoreWeave is one example, and there are many others. SpaceX is in this business in their gross Memphis facility. The big providers lease capacity in addition to their own capital investments. It helps them hedge risk. When the merry go round stops on AI, these will all go bankrupt, and H100s will be like a circa 2001 Aeron chair.
- treis 4mo agoThere are things called LLMs that are incredibly useful but require a staggering amount of compute. Providers are building out data centers to meet the new need.
- linsomniac 4mo agoIn the "AI 2027" article/story, they're talking about next generation models "using 100 times the compute used to train GPT4. Here's a video covering that part of it https://youtu.be/5KVDDfAkRgc?si=TR8XDQmye6O6kkiT&t=169 https://youtu.be/5KVDDfAkRgc?si=TR8XDQmye6O6kkiT&t=169 (the 100x is at 4:38, this video starts where he starts talking compute of GPT3). So my guess is that what is driving the need is that (speculated) 100x the next major models.