[fr4nkr]

The major data centers being built for AI are much more energy-hungry than car manufacturing, and they're being built at a pace that the US energy grid simply cannot accommodate in the short term... or quite possibly even the long term, considering the US's extreme aversion to expanding nuclear power.

Also, you can call it Luddism if you want, but a car factory is going to bring a lot more net benefit to the average person than an AI data center. Motorized transportation is essential to modern civilization, fancy chat-bots are not.

[efromvt]

I wonder why this doesn’t get us frustrated with the grid, not data centers. Delays on interconnects for renewables and offshore wind both seem pretty self inflicted.

[TheTaytay]

I keep wondering this too. It feels like such a self fulfilling prophecy: don’t build new power plants. Don’t build nuclear. Get mad when the grid can’t keep up…it’s defeatist and anti-growth-of-any-sort through a different lens.

[0cf8612b2e1e]

To be fair, for decades, electricity consumption has been mostly flat. There has not been a need to massively ramp up new generation or distribution. It is only in the last few years that such mega consumers have come online that is requiring new development at a frantic pace.

[fc417fc802]

Not true. Electric vehicles have been threatening to collapse residential grids for quite a few years now. The US hasn't been making the necessary infrastructure investments for a long time. See PG&E for example.

[0cf8612b2e1e]

For something the size of the electrical grid, you can find regional variations, but the national trend is quite clear. One report from a quick search[0]

  Consumption Growth Acceleration: After 14 years of near-stagnant growth (0.1% annually from 2008-2021), US electricity consumption surged 3.0% in 2024, driven by data centers, electric vehicles, and economic recovery, signaling a new era of demand growth.

[0] https://solartechonline.com/blog/how-much-electricity-does-u...

[fc417fc802]

TBF multiple things can be true. A period of stagnation, a failure to perform sufficient upkeep, and a failure to keep up with new demand.

[trueno]

I mean one has to also consider the current political _and_ geopolitical landscape now when it comes to energy needs. And given the current outlook and environments even states are now operating in with federal overreach shutting down offshore wind farm efforts and more, it's not hard to do the calculus that lands you squarely in this reality:

- most grids can't sustain the AI energy demands at the moment

- literally no one could tell you if scaling up with clean/renewable energy sources to meet demand is even going to get greenlit right now. it is straight up gambling to try and give a black and white answer to it.

so to a lot of degrees i absolutely understand why a state might pump the brakes. this is increased pressure on a limited resource that is squeezing _the peoples_ economic circumstances. pump the brakes because no one is talking about how to greenlight it and scale up the right way so it doesn't result in even more financial uncertainty for people that are already financially uncertain. its absolutely not something i would want to give the go ahead on without guarantees that renewable energy is going to be the backbone of the increased energy demand.

[YetAnotherNick]

Also power is not at all a limited resource as many top voted post on HN thinks it to be. Increased demand decreases the price of power not increases it in the long term.

And in any case ban doesn't make any sense. Instead they could charge different for grid electricity usage, and make the datacenter pay for grid expansion when they start building it.

[anon291]

Because we have decided that electrical generation tech ended once China became better at it.

Instead of dealing with that like adults we are throwing a fit instead

[ch4s3]

> AI are much more energy-hungry than car manufacturing

This is not even remotely true[1]. Large auto plants use on the order of 200–250 megawatts of power, which is either on par with a data center or slightly more than a data center depending on the size and how cooling is handled.

[1] https://www.energystar.gov/sites/default/files/tools/Industr...

[tadfisher]

New AI data center builds are being specified in gigawatts, my friend.

[jasonwatkinspdx]

Yup. Here's slides from last year's HotChips on where AI racks are going: https://hc2025.hotchips.org/assets/program/tutorials/HC2025....

The racks rolling out now are in the 100s of KW each, targeting 1 MW per rack as the rough limit for using 400v DC.

The next iteration is go up to 800v DC, riding the coattails of power management components from the EV industry.

[jmyeet]

I suspect uou've misread that document. It is a good document though. It's saying a large parts plant uses ~188,000 MWh, I think per year.

A modern AI data center uses 20-100MW+ of electricity. Those two things aren't the same. 20MW of continuous electricity use (which AI data centers do) translates to 175,000 MWh of electricity per year. That's about the same as a minimum and might be 5+ times more.

This document is only about energy usage so we have to guess what "large" means in terms of employment but 3000 to 7000 seems to the range. Compared to 20-30.

But AI data centers are worse because they actually produce what I call negative jobs. Their currently only value proposition is in laying off people and otherwise suppressing labor costs. All while the residents all pay more for their electricity with the money no longer have because they got laid off.

[fc417fc802]

> A modern AI data center uses 20-100MW+ of electricity.

I understand the high end builds to have exceeded 100 kW per rack at this point, with the largest sites exceeding 1 GW (ie 10x your upper bound). So the smallest datacenters use as much as the largest auto plants, and the largest datacenters use 100x that.

[bnewbold]

Where are you getting the 200 megawatt number from?

The document you linked says that a large auto assembly plant consumes around 188,000 MWh annually (with regional variation). By my quick math that is less than 22 megawatts baseline load (24/7/365).

There is a mention that natural gas and other fuels being used on-site, are you converting those to MWh equivalent? I'm not as familiar with that conversion, but from a quick online calculator I found it would still be under 75 megawatt for electrical and fuel-equivalent combined.

[bigfatkitten]

And a data centre brings absolutely no economic benefit whatsoever to its local community, unlike something like an auto plant that creates entirely new supporting industries to support it and its employees.

[cheriot]

Property taxes come to mind.

[xphos]

Local municipalities collect this and often get tricked into not collecting it via agreements to host it in or near their town for multiple year agreements. Also the assessed value of the property may not come anywhere near the costs of increased electricity demand, water usage and noise pollution problems. For locals

Their is typically high paying jobs in factories but these places dont employ a large staff beyond construction. It a tough spaces.

[toast0]

Property tax is often assessed on capital equipment owned by a business as well as real estate. So while the real estate may be of limited value, the hosted machines could drive significant tax revenue as AI machines are spendy.

You're right though, that municipalities often waive taxes to woo big projects.

[xphos]

Yeah but i think server farms are tricky to access because they have accountants to argue about depreciating CPUs, second hand CPU deppreication is enormous so its possible to wiggle the numbers down. Not to zero but the challenge is that you have to recoup not just hthe land value but the value related all the infrastructure and resource drain this type of business does which is very non-trivial

[bigfatkitten]

Outside of colo facilities in metro areas, minimising property tax is always a key consideration in choosing a location.

[andsoitis]

> data centre brings absolutely no economic benefit whatsoever to its local community

Following that logic, are you suggesting that data centers should not be built at all?

[bigfatkitten]

I’m saying that communities are absolutely within their right to oppose developments that make their homes and their lives worse for no benefit.

[mediaman]

Loudoun County in Virginia generates $1 billion in property tax revenues from data centers.

It funds half of all of their expenditures.

Can you imagine having half of your total municipal government budget being paid for by data centers?

Their citizens pay much lower property tax rates, and get much better schooling and police.

Henrico County (also VA) took $60 million in unexpected new revenues from data centers and created an affordable housing trust that is subsidizing low-cost housing.

Although these counties are figuring it out, it's an incredible failure in imagination for many of these liberals in other states to look at an immense source of new funding that could support schools, housing and health and just spurn it because they heard from a friend of a friend that they consume a lot of water based on a discredited book with elementary math errors.

[no_wizard]

They’re an anomaly that benefits from a number of factors like being close to the government for contracting, early data centers built there and they tended to congregate and dumb luck.

They’re an outlier and don’t really prove much of anything.

Oregon has lots and lots of data centers and not much to show for it on any front, other than higher electric prices for consumers

[mediaman]

Oregon gave a lot of time-limited property tax breaks. They also don't have a sales tax.

So I would agree that giving away the #1 way that data centers contribute to the government isn't optimal, though you could argue it's a long-term play.

As the tax break terms expire, Oregon will get $450 million in annual property taxes from the data centers, or about 1.4% of the state budget.

[bombcar]

Hopefully they don't end up with a "Digital Detroit" when datacenters start closing.

Though even if the AI market collapses, the capital spent means they'd probably keep operating; paying for 30 employees is much different than paying for 3,000 at a factory. But the datacenter might be owned by the creditors at that time.

[fc417fc802]

> that they consume a lot of water based on a discredited book with elementary math errors.

How exactly do you think they dissipate the heat of a continuous 100 MW or 1 GW power draw? I have no idea what book you're referring to but you can do the math yourself it's quite straightforward.

[valleyer]

The book is "Empire of AI". This blog post explains it pretty well: https://blog.andymasley.com/p/empire-of-ai-is-wildly-mislead...

Basically, the author (of the book) compares a data center outside Santiago to usage of water by humans, erroneously imputing that the average human uses only 200 cc of water per day.

[trollbridge]

Perhaps part of the problem here is that most towns that have proposals for AI data centres (including my own) have the developers demanding 10 year tax abatements, so we aren't going to see any of that tax revenue.

[orenlindsey]

I would argue that with the rise of coding and debugging agents, the AI data centers provide (or will in the near future) even more benefit than a car factory, in terms of digital infrastructure. These technologies are just a lot more invisible so we don't realize how important they are.

[pesus]

Can you explain how important they are? So far the benefits seem to be limited to faster code generation, which doesn't solve any actual problem people were facing, and is greatly outweighed by the negatives.

[pojzon]

Yet we read everyday that Agents generating astronomous amounts of slop and pointless projects are also straining global digital infrastructure.

Which is also “invisible”. Using this technology to make advancements in healthcare is 1% of its usage. While 99% is garbage apps noone needs, memes, deep fake videos and porn.

AI as a whole for now is a net negative for the world.

[mcmcmc]

Don’t forget the portion that’s used for mass surveillance, scams, and other blackhat shenanigans. Or supercharged personalized advertising with dynamic pricing. And you can’t miss propaganda and dark money influence campaigns.

[rurp]

LLMs are and will be used as malware, propaganda, and slop generation agents more than they will be used as debugging agents. The amount of energy that we'll need to consume going forward just to defend against malicious users and to filter down the flood of slop is absolutely eye watering and will continue to grow as far as we can tell.