News for hackers. Stuff that matters.
Re:Just 400 US Data Centers Use More Electricity Than Michigan by renehsz on Monday November 16, 2026 @02:02PM (#50521482)
[delayed]
Re:Just 400 US Data Centers Use More Electricity Than Michigan by WarmWash on Monday November 16, 2026 @02:35PM (#50521950)
>Datacenter power consumption wouldn't be an issue if...

...they didn't use them for AI.

The panic over DC's has nothing to do with DC's. It's entirely about fear around AI and knee-capping AI progress. It's like when conservatives suddenly care about native grasshoppers when they find out a solar farm is gonna be put on the land. You're totally whiffing if you think they actually care about the grasshoppers.

Re:Just 400 US Data Centers Use More Electricity Than Michigan by dweinus on Monday November 16, 2026 @03:15PM (#50522544)
That may be a factor, but there are other, very rational complaints. They are enormous, noisy eyesores. They result in raised electrical rates for anyone living in the region. They are another user of water in areas that may already be stressed for water. They are also setting us up for massive e-waste in 5 years. Those are real problems that you can't just hand-wave away.
Re:Just 400 US Data Centers Use More Electricity Than Michigan by WarmWash on Monday November 16, 2026 @04:10PM (#50523325)
You can't hand wave them away, but you can dissolve them with even a modicum of grounded research. All industry is big noisy buildings, DCs aren't particularly loud either (and don't need to be), electric rates come mostly from distribution costs, not generation costs, DCs use about 2 golf courses of water a year, so ask the farmers to water one less day (and save the ag runoff from hitting the streams too), inference is currently being served on 5 year old hardware and 10 year old cards are even hard and expensive to get.

And last of all, it's like 3 datacenters that keep coming up as "DCs are bad" despite their being thousands in the country (and hundreds of hyperscale sized ones).

The whole thing is public panic rather than academic honesty.

Re:Just 400 US Data Centers Use More Electricity Than Michigan by dweinus on Monday November 16, 2026 @05:39PM (#50524485)
You are telling me I'm wrong, not because any of these impacts are wrong, but that they shouldn't count. That's hand-waving.

> All industry is big noisy buildings But they produce jobs, goods, and social value. If people don't see the value of AI (or even see it as a job risk), then why would they be inclined to put up with the noise and land use?

> electric rates come mostly from distribution costs Which are costs that result in higher residential rates. Go ask someone who recently looked at their ComEd bill.

> DCs use about 2 golf courses of water a year, so ask the farmers to water one less day Except you can't ask farmers to do that. We need food, and water is finite. That's not anti-AI panic, it's a cold fact.

Re:Just 400 US Data Centers Use More Electricity Than Michigan by WarmWash on Monday November 16, 2026 @06:23PM (#50525034)
The tax revenue from datacenters is at the top of "dollars per acre". States given incentives, but towns sure as hell take the payouts.

DCs aren't distribution networks. They pay for their own grid connection.

We don't need to farm in water stricken areas to feed people. Protests against this long predate DCs.

Again, don't fall for the hysteria.

Re:Just 400 US Data Centers Use More Electricity Than Michigan by hrncn on Monday November 16, 2026 @03:34PM (#50522840)
Gas has the advantage of being locally available, without having to tax the grid even more, which is not even keeping up a little bit at this point, and has a decade+ build out backlog. Gas is just very convenient, compared to other options. And that is the real problem, because there is no way to fix this quickly.
Re:Just 400 US Data Centers Use More Electricity Than Michigan by AnthonyMouse on Monday November 16, 2026 @04:28PM (#50523552)
> A sustainable hyperscale datacenter can be built. There just isn't enough of an incentive as long as buying gas and polluting the air is cheaper (because the cost of both is externalized on society)

This is a misidentification of the problem. Generating power on-site with turbines is an alternative to bidding up the price of grid power that raises the short-term price of electricity for other consumers. If they're doing one they're not doing the other.

Moreover, small gas turbines and converted aircraft engines aren't actually that the thing with the lowest long-term operating costs, they're just the thing you can plop out of a truck and turn on right away.

Which means the best thing might be to use turbines temporarily to avoid raising short-term electricity prices for ordinary people, but only for the time it takes to build some new non-carbon generation to replace them.

And the latter is what everybody really wants, because renewable energy isn't just cleaner, it's cheaper. It just takes longer to deploy, not least because the permits cost more and take longer than they ought to. Which means that if you really want to get rid of the turbines and the high electric bills, what you need is to make it faster to build new non-carbon generation.

Re:Just 400 US Data Centers Use More Electricity Than Michigan by renehsz on Tuesday November 17, 2026 @11:25AM (#50533765)
> Generating power on-site with turbines is an alternative to bidding up the price of grid power that raises the short-term price of electricity for other consumers. If they're doing one they're not doing the other.

They could also do neither and build carbon-free generation (and storage and transmission that goes along with it) from the get-go. That's what I'm arguing for. Not simply making grid operators deal with it.

I'd love to see some of that insane AI money be used to make actual progress towards net-zero emissions now.

> the best thing might be to use turbines temporarily to avoid raising short-term electricity prices for ordinary people, but only for the time it takes to build some new non-carbon generation to replace them

My fear is that some of these gas turbines aren't as temporary as they want people to believe. But even if we take their word for it, why spend money on a temporary solution and on a permanent solution later, even though we have the technology and supply chain to build the permanent solution now?

> renewable energy isn't just cleaner, it's cheaper

I agree that it's fundamentally cheaper, though some politicians seem to be dead set on skewing the market unfairly towards fossil fuels.

> It just takes longer to deploy, not least because the permits cost more and take longer than they ought to.

And I agree that delays due to permits are problematic, but I'd love to see AI money and momentum being used to overcome these hurdles. It shouldn't take longer from a technical perspective.

Right now, the voting public largely opposes datacenters being built at all, and that can hardly be the outcome we should strive for.

Instead, we should be building datacenters that are fully sustainable from the get-go and, if anything, have a positive effect on grid prices.

Rather than rejecting datacenter projects outright, let's push them to fulfill that goal for the benefit of all of us. Hundreds of billions of dollars from the AI hype should be enough to do it.

Re:Just 400 US Data Centers Use More Electricity Than Michigan by AnthonyMouse on Tuesday November 17, 2026 @12:39PM (#50535105)
> But even if we take their word for it, why spend money on a temporary solution and on a permanent solution later, even though we have the technology and supply chain to build the permanent solution now?

Then you need the permanent solution to be able to come online immediately so that you don't need a temporary solution, and that's the thing being held up by permits etc.

> Not simply making grid operators deal with it.

Dealing with it is literally the purpose of a grid operator. The place you want the consumer and the place you want the generator aren't necessarily the same place, which is the purpose of the grid. The grid operator then adds capacity according to expected demand. The long-term effect of higher demand tends to drive prices down because it improves economies of scale.

The short term problem is that the grid operator is subject to the same permitting delays etc. as anyone else, and then if demand increases it takes too long for supply to respond and in the meantime you get high prices.

> And I agree that delays due to permits are problematic, but I'd love to see AI money and momentum being used to overcome these hurdles.

This isn't so much a money problem as a policy problem. Renewable energy projects require pre-approval from NIMBYs when they shouldn't. To solve that you need to change the law. It's not a money problem unless you intend to use the money for bribery.