I guess there are no laws on generating child porn in space which is why he is so desperate for it. Can't see any other reason to put it in space, so pointlessly expensive. Sticking it underwater would be easier.
Fuck AI
"We did it, Patrick! We made a technological breakthrough!"
A place for all those who loathe AI to discuss things, post articles, and ridicule the AI hype. Proud supporter of working people. And proud booer of SXSW 2024.
AI, in this case, refers to LLMs, GPT technology, and anything listed as "AI" meant to increase market valuations.
Rules:
- Relevant posts only.
- No bigotry.
- No misinformation.
- No NSFW/NSFL content without being labeled NSFW or being in an appropriately-labeled spoiler tag.
- This isn't a place to talk about the cons and (supposed) pros of AI for different use cases. This is /c/fuck_ai .
- Especially no AI advocacy.
- No AI-generated content. If you want to post examples of AI failing hillariously, there are other communities for that like !ArtificialIgnorance@lemmy.ca .
- Venting is part of the purpose of this community, but don't aim frustration or accusations at other users.
- These rules are a living document that may change in the future.
You can help the mod team by making your reports detailed.
Effective reports:
- "The third image is a well-known anti-semitic trope"
- "Claiming that 'agentic AI' can outperform humans at all jobs is misinformation"
Less effective reports:
- "Breaks rules"
- "Misinformation"
We kept doing things that will eventually lead to an environmental catastrophe. Do you think billionaires would care about that?
yes because it'll lead to more junk they can sell to the masses..
This is such a recurring pattern with Elon and I feel like even discussing it to dunk on how technically infeasible it is is just part of his grift. Every time it’s these impossible goals that get idiot investors hyped, and pointing out the impossibility just keeps it in the news.
Apart from the issue that it would not really work. And there are a number of reasons.
First, they do have cooling issues down here on earth already, where there is water and atmosphere. Out in space, you can only cool by radiation. That's why the ISS has those gargantuan cooling fins (and no, idon't mean the solar panels, look at the other side, basically the "rear"). And they are relatively economic with their heat generation up there, not like a data center.
Second, none of the current high performance AI chips are radiation hardened. Nor are the memory chips, or any other electronic components they use. In spaceflight, they are using rather outdated processors in their radiation-hardened version for reasons. Anything modern high performance stuff would throw errors (best case) or fry (not so best case) without serious shielding.
Nevermind the fact getting it there is expensive and these products are expected tonbe obsolete on a couple years anyway, at least on earth many of the metals can be recycled but I expect this would just create more space junk as they decide it's cheaper to throw it away. Even sticking it in landfill means it's on earth and technically accessible sfill. These resources will be effectively destroyed.
No one wants to pay for it, either.
Since the debt for data center projects is mostly collateralized on the value of the GPUs, SpaceX's lenders are refusing to lend them money for assets they plan to launch into space — the rationale being that if SpaceX can't pay back the billions in loans, there will be no way to recover the collateral from orbit.
https://shattered.io/spacex-lenders-ai-chips-space-launch-ban-43-billion-project-saturn/
I don't believe you are wrong about using the chips as collateral but even doing that on Earth is stupid. These chips don't appreciate (ignoring the 5080 bs). The moment they are created they start to lose future value.
This is grade 10 flunkee science.
This is probably never going to happen because the vacuum of space is an exceptional insulator. Computers need to dissipate a lot of heat, and cooling anything in space is really hard. You can only transfer heat through radiation in space, and that is way less efficient than convection (the method of heat transfer we use to cool computers on Earth for the most part; there are some parts that use conduction). Cooling computers on space is harder than cooling them on Earth.
On top of that, chips will have to be shielded from radiation interference from the Sun and deep space sources. The magnetosphere protects us from that on Earth, but there is not as much protection in Earth's orbit. This is the reason why any chip we send to space is a few generations behind anything we use on Earth.
Finally, you have to launch everything into space, which is waaaaaaaaaaay more expensive than shipping anything across the world. I short you end up with a hard/expensive to cool data center, full of old chips, that you paid too much to build. It makes no sense.
The only upside that I can see here is that Elon owns the companies that would build all of this. So he can benefit from investor money and government grants, even if the product is shit or it doesn't work at all. I'd bet this is all a scam.
A funny twist thats happening with these is that often times the bonds used to build datacenters use the GPUs as collateral if the datacenter fails.
News has come out that SpaceX is having trouble financing these orbital datacenters because the lenders are saying they won't be able to access or resell the GPUs, so they won't accept them as collatoral for the loans.
Physics is the main problem with these, but the financials are running a close second.
"One million data centers in space!"
-Some childish drug addled loser.
Hyperloop all over again. At most they'll launch one trial satellite.
I don't get why people are even taking it seriously. It's not going to happen because it doesn't make economical sense. The hardware they want in these things has a lifetime of 2.5 to 3 years, and you can just replace it once it's in space.
It's a stupid idea even before you get to the engineering part of it.
It would also be rediculously inefficient and stupid, since heat doesn't simply disappear in space. Sure, space is "cold", but it's cold because of a lack of matter to store heat. You're going to have a very hard time transfering heat to matter that isn't there. "just radiate it", sure, good luck with radiating heat from a whole datacenter.
The whole idea is fucking moronic. It's made up by someone who thinks he's a genius, but who apparently failed 7th grade physics. It's a childish tantrum, "well if you won't let me build my datacenters on earth, I'll just fly them to space!!!"
Unless SpaceX somehow figures out photonic processors, which world produce almost no heat and would be speedy and efficient, in time before AI investment dries up, it's not feasible at all, just like Musk promising that there will be a self sustaining Martian city of 1 million people by 2050, not likely at all.
Or a fully self-driving car in 2016.
Or a hyperloop connecting cities.
Or that DOGE will save 2 trillion dollars.
He's a con man.
not the dyson sphere we asked for
Best we can do is a Kessler.
looks like there is no way out of this one
why would they put into space? its really hard to radiate those heat away.
didnt google tried to put it in the ocean and was tremendous success? what happen to that
Engineering Rule #1: Everything leaks.
I think that was Microsoft and AFAIK it did work but maintenance costs were high.
Yes, let’s let Elon boil the oceans with a million data centers
Because starship needs something to do.
It’ll just explode on the way there as usual. The debris is the real environmental catastrophe.
"You had me at environmental catastrophe! Let's goooo!"
-- Conservatives
The ketamine addict opened his mouth yet again to spout some dumb fantasies that will never happen, I see?
I'm assuming that again the fanboys and investors are showering him with money over his dumb bullshit THAT NEVER HAPPENS?
I dont get it. If I say dumb shit, people laugh. If he says obvious completely nonsensical bullshit people invest literally billions over billions of dollars (hyper loop comes to mind). What am I doing wrong?