So, re: LLMs and the "jury is out on energy usage", I just want to point out a couple of historical things.

When studies came out linking smoking to increased risk of cancer, the Tobacco industry paid for and produced studies that said "nuh uh".

When Obesity and heart disease rates increased in the United States and the problem was linked to sugar intake, the sugar industry produced studies saying it was fats.

https://pmc.ncbi.nlm.nih.gov/articles/PMC3490543/#sec4

Inventing Conflicts of Interest: A History of Tobacco Industry Tactics

Confronted by compelling peer-reviewed scientific evidence of the harms of smoking, the tobacco industry, beginning in the 1950s, used sophisticated public relations approaches to undermine and distort the emerging science. The industry campaign ...

PubMed Central (PMC)

The fact that the GenAI industry isn't providing actual data about their energy usage should make you _very_ skeptical about their claims of "it's fine".

And I'm a little disappointed in seeing folks I generally respect being like "oh no, it's not as much as you think" - y'all need to show me hard data.

Also, I would fucking hope a text extruder uses less energy than serving 4k video, the fact that it presently *isn't* is a travesty - nevermind its other travesties.

@cthos

The fact that the GenAI industry isn't providing actual data about their energy usage should make you very skeptical about their claims of "it's fine".my brain is just screaming "this this this", because the fact that they don't provide data while simultaneously pursuing access to larger and larger power sources is a huge red flag. If their claims were true, the data would be published, proudly. They would brag about it.

This isn't taking into account their partially externalized power costs for training and such, either; scrapers are massively increasing compute demands on other servers, which translates to higher power costs that they have no way of quantifying (and are definitely not interested in doing so).

@aud "Yeah, we totally need nuclear power but we pinkie swear we're making stuff more efficient. Oh look, new more power-hungry Nvidia cards!"

@cthos @aud Well, considering that nuclear power is one of the most expensive power sources, that kind of makes no sense.

Literally, even with massive state subsidies, they are still the most expensive per MWh. Without massive state subsidies, nobody would even consider building these crazy things. And if you factor in the costs of getting rid of the nuclear waste properly, the correct costs raise off the graph, even with a log scale.

So not exactly sure what the angle here is.

@yacc143 @aud Correct, their claims about energy usage (when they bother to make claims) don't make sense. ^_^

To be accurate though I've only seen them trying to recommission existing reactors rather than building new ones so I was being *slightly* hyperbolic above but, wouldn't put it past them.

And if we look at Musk's xAI plant I think they just intend to pollute until someone makes them stop, so.....

@cthos @aud Ah, funny that, they probably want these new "advanced nuclear plants", that are supposed to be cheaper, but the only existing such plant, Plant Vogle, GA, 3&4 is estimated at LCOE $100-180/MWh while the advertised “prospectus” values for advanced nuclear plants are in $80-$85/MWh range.

Note that "analysts" (who partially admit that they don't consider all the financing costs) are on the lower end of the bracket, while the Princeton researcher gets a whopping $178/MWh.

For comparison, wind/PV are in the $30/MWh range, and you can locate data centres in consistent weather zones. With that energy usage, batteries, and reserve capacities literally pay for themselves (Hint, hint: Texas, the guys with their private grid, and thus grid balancing issues from hell, have been literally subsidising crypto miners to have a huge electricity user, that they can basically turn off instantly, if they need to balance the grid.)

With commercial financing btw, old style nuclear plants, when newly built have LCOEs slightly under $200/MWh

Even in the USA, these things are literally the most expensive form of generating electricity.

Ah, I'm more relaxed. At some point, somewhere the costs will have to meet the usefulness.

Yes, these things have a certain usefulness (especially when harnessed in higher level algorithms that keep them on track.)

But the question is, will the providers of these model be able to charge what it costs them (and make a somewhat relevant profit, capitalism lives for the profit).

And almost all applications that produce at least somewhat useful results use big contexts, which implies costs.

@cthos a bit of hard data from a point of view AI companies don't want to talk about: @iocaine

Daily stats from my crawler defense system, showing 7-8 million requests a day, against the very small and niche sites I host.

98% of my traffic is crawlers. It is their cost, which they try to push onto others. It's completely ridiculuous.

Even if the energy use of training and generating would be zero, the crawling cost, paid by literally everyone else, is not. This cost is easily demonstratable.

@algernon @iocaine Oh yeah, externalized energy usage is a whole thing too (as @aud mentioned in a reply)! Thanks for that.

Relatedly, the "these things use a lot of energy" crowd has shown their work in several articles, but the rebuttals are always along the lines of "your assumptions are wrong" backed up with 'vibes'.

@cthos @algernon @iocaine @aud One article I've read squared the estimate it was citing for LLM energy use against how much these companies are demanding for their new server farms by saying:

"Its the other forms of AI which uses all that energy!"

Which I find to be an ahistorical stance to take!

@cthos Sure all those GPU's Nvidia sold in the recent years are just sitting idle in some basement...
@cthos to say nothing of what fossil fuel companies did

@Yza Yeahhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh.

That would have made for a more direct comparison, wouldn't it?

@cthos that's the exciting thing about capitlaism! there are always more examples
@cthos well in [some cases](https://naacp.org/articles/elon-musks-xai-threatened-lawsuit-over-air-pollution-memphis-data-center-filed-behalf) it doesn’t matter. It’s so bad they’re jumping straight to absurd emissions causing localized air quality issues.
Elon Musk’s xAI threatened with lawsuit over air pollution from Memphis data center, filed on behalf of NAACP

On Tuesday, the Southern Environmental Law Center, on behalf of the NAACP, sent a letter to xAI signaling their intent to sue over the company’s continued use of unpermitted gas turbines at its data center in South Memphis.

NAACP

@elebertus Yeah, I mentioned that one in another reply downthread, that one is the most egregious but if they think they can get away with it, well....

Edit: "The most egregious in this particular moment in time that I am aware of", I'm sure there's more.

@cthos absolutely an outlier since their general business practice is to do what they want until forced not to.
@cthos maybe they mean the jury can't work because of the power cuts
@cthos
TBH, "the jury is out" over whether LLM energy usage is just bad or extremely terrible. Either way it should be cut...