Live data from Hacker News

Stargate Project: SoftBank, OpenAI, Oracle, MGX to build data centers

apnews.com

901–910 of 1001 posts

Re: Stargate Project: SoftBank, OpenAI, Oracle, MGX to build data centers

#901
post #351

Earlier quoted context omitted.

I'll make a wild guess that they will be building data centers and maybe robotic labs. They are starting with 100B of committed by mostly Softbank, but probably not transacted yet, money. > building new AI infrastructure for OpenAI in the United States The carrot is probably something like - we will build enough compute to make a supper intelligence that will solve all the problems, ???, profit.

If we look at the processing requirements in nature, I think that the main trend in AI going forward is going to be doing more with less, not doing less with more, as the current scaling is going. Thermodynamic neural networks may also basically turn everything on its ear, especially if we figure out how to scale them like NAND flash. If anything, I would estimate that this is a space-race type effort to “win” the AI…

What is a thermodynamic well? Couldn't find much on it.

Re: Stargate Project: SoftBank, OpenAI, Oracle, MGX to build data centers

#902
How much exaFLOPS can we expect from a 100 Billions dollars datacenter today? A rough estimate from a quick Perplexity search gives us 24 exaFLOPS for all smartphones in the world and 12 exaFLOPS for personal computers. Could a competitor to such a Datacenter be a collective effort with some sort of crypto to split the benefits?

Re: Stargate Project: SoftBank, OpenAI, Oracle, MGX to build data centers

#903

Earlier quoted context omitted.

there doesn't appear to be any timeline announced here. the article says the "initial investment" is expected to be $100bn, but even that doesn't mean $100bn this year. if this is part of softbank's existing plan to invest $100bn in ai over the next four years, then all that's being announced here is that Sama and Larry Ellison wanted to stand on a stage beside trump and remind people about it.

The literal first sentence of the announcement is: > The Stargate Project is a new company which intends to invest $500 billion over the next four years

Not the first sentence of the new AP link target, which is much more vague. Kind of annoying for HN to swap it out like this.

Re: Stargate Project: SoftBank, OpenAI, Oracle, MGX to build data centers

#904

Earlier quoted context omitted.

If they're creating that many jobs, it means most of them are construction work. Skilled labor for sure, but not necessarily college educated.

How does this work out in the long term? Operating a data center does not require that many blue-collar workers. I'm imagining a future where the US builds a Tower of Babel from thousands of data centers just to keep people employed and occupied. Maybe also add in some paperclip factories¹? ¹) https://www.decisionproblem.com/paperclips/index2.html

It's just lies they use to sell the idea that these profit driven corporations should be given massive amounts of money at the expense of the commons and public.

It's no different than sports stadiums selling the same idea to local governments:

"We'll create jobs."

"Mostly low quality jobs with poor benefits and minimum wages."

This is no different.

Whenever you hear about job claims you should be asking what quality of jobs?

Re: Stargate Project: SoftBank, OpenAI, Oracle, MGX to build data centers

#905

I'm confused and a bit disturbed; honestly having a very difficult time internalizing and processing this information. This announcement is making me wonder if I'm poorly calibrated on the current progress of AI development and the potential path forward. Is the key idea here that current AI development has figured out enough to brute force a path towards AGI? Or I guess the alternative is that they expect to figure…

"There are maybe a few hundred people in the world who viscerally understand what's coming. Most are at DeepMind / OpenAI / Anthropic / X but some are on the outside. You have to be able to forecast the aggregate effect of rapid algorithmic improvement, aggressive investment in building RL environments for iterative self-improvement, and many tens of billions already committed to building data centers. Either we're a…

I am not qualified to make any assumptions but I do wonder if a massive investment into computing infrastructure serves national security purposes beyond AI. Like building subway stations that also happen to serve as bomb shelters.

Are there computing and cryptography problems that the infrastructure could be (publicly or quietly) reallocated to address if the United States found itself in a conflict? Any cryptographers here have a thought on whether hundreds of thousands of GPUs turned on a single cryptographic key would yield any value?

Re: Stargate Project: SoftBank, OpenAI, Oracle, MGX to build data centers

#906
post #589

This is so much money with which we could actually solve problems in the world. maybe even stop wars which break out because of scarcity issues. maybe i am getting to old or to friendly to humans, but it's staggering to me how the priorities are for such things.

It's an indirect attempt of tackling any first order problem. So is all software engineering.

Re: Stargate Project: SoftBank, OpenAI, Oracle, MGX to build data centers

#908
post #902

How much exaFLOPS can we expect from a 100 Billions dollars datacenter today? A rough estimate from a quick Perplexity search gives us 24 exaFLOPS for all smartphones in the world and 12 exaFLOPS for personal computers. Could a competitor to such a Datacenter be a collective effort with some sort of crypto to split the benefits?

apparently around 100 exaflops if the 100 billions dollars datacenter is made only of nvidia H100. I am guessing the fact that (most?) personal computers and smartphones do not have a big gpu makes the above scenario difficult.

Re: Stargate Project: SoftBank, OpenAI, Oracle, MGX to build data centers

#909
post #589

This is so much money with which we could actually solve problems in the world. maybe even stop wars which break out because of scarcity issues. maybe i am getting to old or to friendly to humans, but it's staggering to me how the priorities are for such things.

I am surprised at the negativity from HN. Their clear goal is to build superintelligence. Listen to any of the interviews with Altman, Demis Hassabis, or Dario Amodei (Anthropic) on the purpose of this. They discuss the roadmaps to unlimited energy, curing disease, farming innovations to feed billions, permanent solutions to climate change, and more. Does no one on HN believe in this anymore? Isn't this tech startup…

Just my opinion/observation really but I believe its because people are implicitly entertaining the possibility that it is no longer about software or rather this announcement implicitly states that talent long term isn't the main advantage but instead hardware, compute, etc and most importantly the wealth and connections to gain access to large sums of capital. AI will enable capital/wealthy elite to have more of an advantage over human intelligence/ingenuity which I think is not typically what most hacker/tech forums are about.

For example it isn't what you can do tinkering in your home/garage anymore; or what algorithm you can crack with your intrinsic worth to create more use cases and possibilities - but capital, relationships, hardware and politics. A recent article that went around, and many others are believing capital and wealth will matter more and make "talent" obsolete in the world of AI - this large figure in this article just adds money to that hypothesis.

All this means the big get bigger. It isn't about startup's/grinding hard/working hard/being smarter/etc which means it isn't really meritocratic. This creates an uneven playing field that is quite different than previous software technology phases where the gains/access to the gains has been more distributed/democratized and mostly accessible to the talented/hard working (e.g. the risk taking startup entrepreneur with coding skills and a love of tech).

In some ways it is kind of the opposite of the indy hacker stereotype who ironically is probably one of the biggest losers in the new AI world. In the new world what matters is wealth/ownership of capital, relationships, politics, land, resources and other physical/social assets. In the new AI world scammers, PR people, salespeople, politicians, ultra wealthy with power etc thrive and nepotism/connections are the main advantage. You don't just see this in AI btw (e.g. recent meme coins seen as better path to wealth than working due to weak link to power figure), but AI like any tech amplifies the capability of people with power especially if by definition the powerful don't need to be smart/need other smart people to yield it unlike other tech in the past.

They needed smart people in the past; we may be approaching a world where the smart people make themselves as a whole redundant. I can understand why a place like this doesn't want that to succeed, even if the world's resources are being channeled to that end. Time will tell.

Re: Stargate Project: SoftBank, OpenAI, Oracle, MGX to build data centers

#910
post #808

Earlier quoted context omitted.

Let me avoid the use of the word AGI here because the term is a little too loaded for me these days. 1) reasoning capabilities in latest models are rapidly approaching superhuman levels and continue to scale with compute. 2) intelligence at a certain level is easier to achieve algorithmically when the hardware improves. There's also a larger path to intelligence and often simpler mechanisms 3) most current generation…

>reasoning capabilities in latest models are rapidly approaching superhuman levels and continue to scale with compute I still have a pretty hard time getting it to tell me how many sisters Alice has. I think this might be a bit optimistic.

They plugged the hole for "how many 'r''s in 'strawberry'", but I just asked it how many "l"s in "lemolade" (spelling intentional) and it told me 1. If you make it close to, but not exactly a word it would be expecting it falls over.
Post reply on HN