They haven't. Excepting RC4, which is wildly different from every other mainstream cipher, the track record of practical block cipher cryptanalysis (of core algorithms) is meager. If you redesigned DES so that it had 16-byte blocks and keys and the round structure to match, but retained the rest of its core design decisions, it'd presumably be viable (if slow).
There's really not much to it? Stick it in a low oven and bring it just under temp, then finish it in a ripping pan. If the results you're getting from that aren't restaurant-caliber, it's probably the meat you're buying. You definitely do not need 550C to put an excellent crust on a steak!
There was an opportunity to organize and exploit market power back in 2017-2018, when there was a frenzy of recruiting and growth in the industry. The field was also much more ideologically coherent, in the wake of the first Trump election, than it is today. But nothing came of it. It's less clear that any similar opportunity exists today, or why it would come together now when it didn't under more favorable circumstances then.
It's inconvenient for any amount of data, because it essentially begs the question; if you can securely transmit N bytes of key pad to a counterparty, just use that mechanism to transmit N bytes of plaintext instead.
It has the advantage that the key can be sent before the message is known. Think military battlefield. Your commander goes out to war with a CD, and then he can transmit messages like "we encountered the enemy". It would do no good to transmit "we encountered the enemy" before the war started.
The funny thing about data center concerns in Chicagoland, where I'm hearing it, is that data centers are better neighbors than almost any other commercial real estate. The data centers in Aurora are surrounded with prairie green space; everything else, including the nicest office buildings, are a sea of eyesore asphalt parking lots.
I get there can be real concerns, locally-specific, about energy usage (again: they depend on where you are). But data centers by themselves intrinsically seem like very good neighbors, much better than warehouses and strip malls.
The reputation of datacenters precludes them due to some high-profile abuses of locals, as well as cities overcommitting brittle and unprepared infrastructure to them.
Consumer energy prices in 13+ states hiked because the grid was unable to meet the sudden demand. SpaceX's datacenter in Memphis used gas turbines to rapidly deploy Colossus before the grid could be upgraded, which resulted in an increase in local air emissions that was technically illegal. There are two Meta datacenters that oneshotted their local water supplies during construction and tests, resulting in sediment and bacteria going through consumers' taps. NV Energy cut off an entire town of residential properties because datacenters were more profitable, in a higher-stakes parallel to the ongoing RAM woes in consumer electronics.
There's been more than enough well-justified association between "datacenter" and "abuses local resources" that I think nobody should really be surprised at the backlash toward them. The backlash is so furious that it's forming a rare example of a bipartisan, populist consensus on any given political issue.
The infrastructure thing I get! The water thing is, I think, baloney, but depending on where you are there's energy generation concerns. Seems totally reasonable to me for local political bodies to demand that new data center projects account for what they'll do to the local power grid, and, obviously, running jet turbine gas generators is not a reasonable answer to those concerns.
What frustrates me about the new political valence of data centers is that there are reasons to be optimistic about data centers in blue states, which could move the market towards renewables and grid modernization, making data centers --- which are going to happen one way or the other --- a vector for improving our current energy system. Instead, they're all just going to be pushed to the locales who will do the least to regulate energy usage.
It doesn't help that at least one datacenter operator and possibly more promised municipalities in Arizona that they wouldn't use their water and then secretly used millions of gallons. This in the middle of water shortages.
I don't know what timeframe we are talking about here, but 1 million gallons is how much 33 people in the US use in a year. It's really not that much water.
> 1 million gallons is how much 33 people in the US use in a year.
For a moment I was worried that this might be an average that includes indirect usage through industry and agriculture, but AFAICT no, it's in line with just residential usage, so good-on-ya for that.
> I don't know what timeframe we are talking about here
The utility claims it was 30m gallons over 4 months [0], and 90mgal/yr would be like ~3k individuals. This seems pretty significant given that Fayetteville is only about ~20k residents. [1]
I've been checking stats and comparing it with here (Spain) and I never thought the difference of residential water usage was so different.
Average use of water per capita in Spain is about 47m3 a year, so a third of the USA. And that number is quite consistent around the EU, except for Greece and Cyprus, where it's about double or even a bit more.
I think it's less the hard numbers than "promised not to use then used". It destroys trust in any future developments because you know you can't take them at their word.
This is the emotional truth that a lot of people are just not seeing.
It takes so long to build trust. It takes seconds to obliterate it. And big tech as a whole has been obliterating it's trust for like... a decade now.
You can't just isolate AI data centers, or even this potentially not-AI one, from everything else and be like "well opposition to this normal thing is ridiculous." Yes, it might well be. However there has been a pattern of abusive behavior on the part of AI DCs ranging from:
* Getting zoning approval seemingly against the will of a preponderance of local citizenry
* Taking water, or in vastly more quantities anyway, without permission
* Tapping water sources they weren't permitted for
* The above two of which, I'm pretty sure were only found out when water pressure started dropping in people's homes
* Similar to the above, energy demands causing residents to be told to turn down their air conditioning, during a heatwave no less
* Resulting in large scale change to the places people live, irrespective of benefits, that they didn't ask for, had no say in, and especially given it's for AI, likely have strong feelings about
* The general vibe lately that big tech is above the law, not just zoning laws but seemingly any laws they believe "stifle innovation"
Like, you can disagree with all of these points, and give rational, logical arguments to them. And most of them I would agree with you! But just because YOU don't think there's a problem, doesn't mean people don't think there's a problem, and most importantly: you cannot decide there isn't a problem for them.
And you can take this even further, expanding beyond just AI into things like the ongoing scandals with Flock cameras, Big Tech's involvement with Israel and it's war, the ongoing scourge that is social media, alongside the old classics: products are broken all the time, services are shut down with little warning, the price of every piece of tech and software seems to constantly be rising, many of which have become more or less essential to life, and at the top of all of these companies are, reliably, the richest people on the damn planet, who seemingly are only getting richer, all the time. TWO of them have their own SPACE programs for goodness sakes.
Like I'm sure the majority of people here especially see these things as progress, even aspirational, or at worst, a cost of doing business. What a lot of the public sees is a massive cohort of bullies and oligarchs, seemingly accountable to no one, who are so rich that they can simply ignore laws they deem beneath their ambitions with the pleased signoff of their government. That's not a good situation and simply saying "well they're just dumb luddites" is not a solution.
This shit is unstable.
People don't see tech as a benefit anymore, not remotely the way they used to anyway. It feels more and more like any time these companies make their presence known in your life, it's a threat to you, or things you care about. And if these companies don't start taking that seriously and try to mend some bridges, I think we're about to see the temperature really start to rise.
I can't get past the premise here that data centers are using material amounts of water. You took all the time to write this, but then included an assertion I think we all broadly know isn't true.
I think you’re completely sidestepping the point of the OPs post about “emotional truth”.
It’s indisputably true that a data center tapped a water supply without the supplier knowing about it and it caused water pressure in surrounding homes to drop:
Whether it was a “material amount” of water is pretty irrelevant here. “affects my home water supply and they didn’t pay for it” makes it plenty material for voters and just feeds into the existing narrative that these companies are stomping all over the will of everyday people.
I don’t really get the point you’re making. That it’s illogical? Sure. Voters often are. They still get to vote, though. Sometimes you have to sell an idea through emotion. “I assure you the data center is not using a material amount of water” is not going move the dial.
> You can disagree with all of these points, and give rational, logical arguments to them. And most of them I would agree with you! But just because YOU don't think there's a problem, doesn't mean people don't think there's a problem, and most importantly: you cannot decide there isn't a problem for them.
And, really, I don't think it's that their water usage is uniquely terrible, I think it's far more that their water usage feels unnecessary which goes back to the AI not doing much for regular people issue. Most people see AI features as intrusive nuisances, and even in the best of circumstances, a huge AI DC is going to need a lot of water, both directly for cooling, and indirectly via it's demand to the power grid. Virtually every method aside of solar we use to generate power requires pretty substantial amounts of water, if for nothing else than to boil it.
That doesn't mean that data center is using a huge, unbelievable amount of water, it's that whatever water it is using, which we know is at least substantial, is water that could be better used for other things. Nobody bats an eye at the amount of water used to make concrete or grow food or raise livestock, because we have to get around and build things, and we have to eat. But if you have this tech which is seemingly a lightning rod for nonsense that needs it's own special data centers that use tons of the stuff... people are gonna say something.
Why sweep for them? The industry promises/makes commitments with the intention of not keeping them and going through back channels. F em. They are a scumbag/untrustworthy/'we only technically broke the law' industry at this point.
Just like everything else, it depends! There are designs that don't really use that much water, but there are also ways to build these things that absolutely destroy the local water supply. It's a legitimate concern that shouldn't be a deal killer without any discussion, but it also should be discussed and considered and understood how any given project is planning to use water
I know of one example where this was a problem, it didn't destroy the local water supply but water usage was absolutely a problem. A family friend is a sales engineer for an industrial/commercial HVAC manufacturer and they had quoted a chiller that I think was basically a swamp cooler, in that it would constantly need new water as it cooled because it was using evaporative cooling
Turns out they could not get enough water from the local water utility and a ground water source was not clean enough so they had to build a water pipeline all the way to a neighboring town to supply water for the chillers
This is what I was talking about, depending on the HVAC setup it may be closed loop but also it could be evaporative swamp coolers that just absolutely destroy a ton of fresh water daily
> What frustrates me about the new political valence of data centers is that there are reasons to be optimistic about data centers in blue states, which could move the market towards renewables and grid modernization
Companies often want to put data centers in the rural parts of blue states (like the southwest of Minnesota, where I live), where the locals are much more red and skeptical, if not outright hostile, toward renewables.
It’s legal/permitted in most cases, but it’s certainly a problem, what these specific data centers are dispensing into the water table in certain places.
The water sediment thing had nothing to do with the building being a data center I think? Any form of construction would interfere with people drinking well water next to it
Isn't that enough to rethink the construction - it takes a while to build a data center, and I would not be happy that I have to suffer with shitty water because a data center is being built for years.
They’re buying up undeveloped land that would not have been developed for any other reason. If the only developer is a problem, then it doesn’t matter if all developers are a problem.
Datacenters (1) use non-potable water during normal operation and (2) return most of the water they use to the water system, instead of consuming it (as opposed to, say, a golf course).
I didn’t get far until the semantics fall apart, you don’t care if it’s consumption or not when you don’t have water in your tap.
And it’s all based on theory, these plants are using systems far under dimensioned for the task, which is never a good thing. The problem with proponents for these behemoths is that they are all about theory and don’t live anywhere near the life they say is so great.
1. So basically the majority of rural people, so you are saying rural Americans should fight against every AI datacenter built in their area if they want to keep access to drinking water. OK.
2. The existing homeowners were there first. Their community has a commitment to ensure that the largest expense they will ever have, their home, is fit for purpose and not impacted by those that come after. If I move next to an airport, can I just have the airport shutdown? No. But a datacenter can just make homes un-inhabitable because 'that's just how it goes'? BS.
Construction (or at least of the type we're talking here, not fracking and tunneling and other things that affect the subsurface water table) interferes with surface water, not ground water.
…and you get to do that because you pay for clean water and sewage treatment. But that doesn’t mean some company can use the public sewers to dispose of as much bacteria-contaminated water as they like without paying for that externality.
0.) It shouldn’t be considered a national panic to ask States and communities to stop permitting this behavior.
1.) All bacteria is far from equal.
2.) Some of this wastewater is not problematic because of bacteria, but metals and other things - and some is being pumped into streams, not water treatment loops.
Unless I'm missing something, the fill-and-flush procedure that Meta's contractors used dumped waste water into the non-potable sewer system, and it never made its way into drinking water. Still room for Meta's contractors to improve. https://www.theguardian.com/us-news/2026/jul/08/meta-datacen...
You’re spreading misinformation. The CAA requirements applies to “stationary gas turbines.” There was an ambiguity as to whether “stationary” applies to mobile gas turbines.
My intent is to highlight that unless you have evidence of corruption, accusing people of bribery is just slander by another name.
I hold our politicians in no high regard, but I do hold myself to some standard in that I won't accuse someone of corrupt conduct unless I actually have evidence.
The point being, it's laughable to slander someone else's morality while you lie to posture your own.
“The party told you to reject the evidence of your eyes and ears. It was their final, most essential command.”
The evidence of corruption is blatant and obvious, a recent notable example has to do with a pool and its failure to be painted appropriately to the tune of $millions.
There are also immeasurable examples of campaign finance law violations (or “Hollywood accounting”) to effectively “legally bribe” politicians.
It’s zero stretch of the imagination that if the politicians are this blatant with what we can see with our own eyes, that there is significantly more sinister activities that happen in the “darkness.”
The corruption of the Federal (and many local) Government has become so blatant and in the open that it has become nearly impossible to trust that it isn’t corrupt to its core.
So sure, stand on a high horse and demand evidence for every statement as society is actively crumbling in front of you.
The problem with “high road” mentality here is that it is not followed by those who are dragging us down, and the bar of needing to provide evidence beyond a reasonable doubt takes such high levels of effort to achieve, that it is impossible to counteract the blatant corruption with the same velocity.
The corollary would be that, in all likelihood, bribes are documented as rare because you are incentivized to ensure that a bribe is never made public. Has Memphis done this before? Is this strange behavior to allow such a construction project to move forward while harming its residents then slow-rolling solutions?
Either way, this is a forum and not Academic Decathlon.
It's public record that Elon Musk personally spent $100M on the Trump campaign. Elon Musk is now using that political influence. While this may not meet the elements of Bribery under US Code, it clears the bar for an opinion on the Orange Site.
But data centers basically take land and resources and generate almost nothing for the local community. The jobs are minimal after the data center is built. It's a giant windowless box.
I would much rather have a strip mall in my community. Strip malls often have local businesses as tenants (so money made by the business then gets spent in the community), and they provide services the local community can interact with and benefit from. They may not win any architectural awards, but I'll take a strip mall over a giant data center any day.
Property Tax revenue, and they are not a drain on services like schools, roads, fire department, police department etc.
I'm watching this debate in a town that i am from and where I live. Our choices are pretty much be swallowed up by growth from neighboring cities and become a bed room community or solar farms and data centers.
The downside with becoming a bedroom community is you risk creating a negative tax base. where the cost of providing services is more than the properties bring with property taxes. I don't know what the right answer is.
Nobody invests a billion dollars in a facility without first lobbying to get an exemption from property taxes. Most states in the US have passed incentives that include sales and property tax exemptions.
Property taxes are ordinarily a percentage of the value of the land and structures because in most cases it has some approximation to services consumed. A building with twice as many condo units might require twice as many police, create twice as much traffic, etc.
A large data center could plausibly be sitting on a piece of land worth a million dollars and then have a billion dollars worth of equipment in it, but still not require any more police officers or traffic lanes etc. than the million dollar piece of land sitting next to it. It might even require less.
It's easy for a town to come out ahead from that even after providing significant tax breaks. When you have something consuming the same unit of services while having an assessed value of 1000x the typical amount, what percentage tax break they would have to get before the government isn't coming out ahead?
You're missing the point. Towns come out ahead when the people living there come out ahead. I'm not arguing with your math, but your math is also a straight line to a town with no people at all being better. Towns are for _people_.
Where do you get the premise that the people living there don't come out ahead?
You have something which, if it moves in, would pay more in property that than it consumes in services. Even at a significant discount to the ordinary property tax rate. If the net benefit after the discount is +$1M/year to the local coffers then the local people are benefiting in that amount, and it's easy to see numbers like that even after offering significant tax incentives.
You're shifting the goalposts from an earlier comment that claimed they would be "all drain". But that's ok, maybe you could compare that money to what a strip mall would provide in tax? I would find that analysis more convincing than just saying "reduced rate therefore bad", and would probably more accurately illuminate why localities make the decisions that they do.
The legislatures are betting that the data centers stick around or bring in jobs. They won't. They think something that expensive must require a ton of jobs to maintain. It doesn't. They need low taxes, cheap land, cheap water, and cheap power.
>Nobody invests a billion dollars in a facility without first lobbying to get an exemption from property taxes. Most states in the US have passed incentives that include sales and property tax exemptions.
is not at all supported as "exactly the truth" by a link that lists 11 states with property tax abatements, most of which are not full exemptions. Am I mistaken?
What's more, having a statutory abatement is only one way to get around property taxes. What you also see are things like a industrial revenue bonds. Those don't require a special law. They just shift the tax burden to the public.
Again, the idea here is that you're incentivizing investment and that will be offset by (a) jobs created by the facility, since labor doesn't get any tax abatement or deferment, (b) they will be bringing goods and service production to the area, and (c) eventually they will pay property taxes. For something like a traditional auto factory, that's a great incentive.
The issue is that data centers have almost no local labor required. A single McDonald's is likely to employ more individuals than a data center will. And the service they provide is essentially independent of the location. It doesn't matter if it's nearby. That provides almost no advantage. And with such a small labor pool and such a large potential tax burden, there may be no economic incentive to remain in the location instead of building a new data center by the time you reach the end of the abatement.
This notion that legislatures are blind to the number of jobs that will be created is contradicted by your previous link. You can scroll through table 2 for this- Alabama expects 20 jobs, Georgia 5-25, Nevada 10, etc. These states don't seem at all confused about how much local labor is required. And this is what they're setting as minimums; they believe that, absent mandate, the data centers would provide fewer jobs than these numbers.
Any source that they're confused about (b)? I believed you about (a) until I read your source- the notion that they're getting misled by construction job numbers was plausible- but I can't imagine how a government would be so fundamentally confused about what data centers are as to believe they'll provide local services or somehow goods.
What seems more likely to me is that the states/counties/cities believe that they will take in enough tax revenue while providing very little back in government resources that throwing in a discount works out just fine economically. I don't think you've made much of a case otherwise.
I work in construction,
- data centres included - in Australia.
From my experience you’re mostly right, however it’s more like, in order of priority: power availability, water availability, data availability, land price. Cost to upgrade upstream power transmission dwarfs all other costs.
I’d note that after construction ongoing jobs are minimal. Specific example: $2B aud datacentre will support 30-60 jobs after completion. And probably 50% of those are low skill/income security,cleaning maintenance etc.
> Specific example: $2B aud datacentre will support 30-60 jobs after completion.
But that's not out of the norm for industrial activity. Like how many people do you think are actually working in an LTL hub or a paper mill or a food product factory at a given time? Probably about the same. You can see the parking lots and count cars on Google Earth.
The towns are estatic because compared to something that generates physical emissions or truck traffic data centers are pretty clean and low risk as long as the noise can be controlled.
One of the ways Californians prevented Walmarts from getting built was to prove how egregious a drain they were going to be. The stormwater runoff alone was breathtaking. Those parking lots and roofs are crazy.
First, Costco has some of the best employee treatment in the industry. Second, we are talking about blocking unwanted neighbors based on anything that will stick. We shouldn't have regulatory capture of socializing the liabilities and privatizing the profits, but we do and we have. Any resistance at all is by definition going to be a matter of coming at the problem sideways, up to and including fighting dirty.
Blocking unwanted neighbours with anything is exactly what NIMBYism is, and you could make similar arguments that housing developers are externalizing the burden of e.g. traffic or school capacity while privatizing the profits. Laws that enable arbitrary suppression of whatever the locals don’t like have caused generational damage via racism and now the housing crisis, and we should get rid of them.
Yep, and my free speech liberties mean that people can say horrible things to each other too. You can't have it both ways and we live in a society with people who have priorities that are misguided in different ways than our priorities are misguided. There's not one single person on this planet who is right all the time. Them's the breaks, wear a helmet.
NIMBY mostly blocks things that don't already have regulatory capture. Blocking things that are is fighting fire with fire.
I'm just addressing the text of your comment of incredulity at Walmart's building footprint. The tone of this one is different so I'll accept that as concession.
Property Tax revenue is surprisingly tiny from data centers in Quebec.
Some how DCs have low assessment for the building, like way lower then their resale value even.
> they are not a drain on services like schools, roads, fire department, police department etc.
Police - definitely not.
Roads on other hand, DCs are hauling new equipment / decommissioned equipment daily.
I'm in fact in favor of DCs, but tax revenue from them is not impressive at all. At least in Quebec.
I have no say in this process. Of course taxes are based on land and building assessed cost. Montreal and suburbs chose to assess condos at a market rate and even higher, while to assess detached house way below the market and assess data centers below the costs it was recently resold.
> The downside with becoming a bedroom community is you risk creating a negative tax base. where the cost of providing services is more than the properties bring with property taxes.
I can affirm that local fire departments are not spending resources preparing for data centre fires in any special way. However, I can attest a recently built DC used political means to dodge aspects of fire codes (mostly related to building road infrastructure for fire truck ingress).
Amazon also secured a 75% reduction in property taxes for 30 years.
Not a data centre expert, but don't they quite famously pump CO2 into the server rooms automatically if the fire alarms go off, even if there are people in the room?
C02 fire suppression is rare in datacenters. That's more of an offshore thing in my experience (I design fire suppression systems for a living, and do life safety systems).
Typically server rooms are currently Fluoro-K / Novec123 if they have space constraints on the suppression systems. Those cylinders take up the least space per cubic foot of protected environments. These are less harmful than previous gasses like Halon, but still not good for the environment.
If you're lucky we get given ample room for cylinders and can go with Argon suppression systems, but that can take up more than 4 or 5 times the space alone and adds a lot of cost in engineering manifolds, etc.
CO2 takes up as much space as Argon systems and will kill someone who doesn't leave the hazard area before release so its generally not used in any datacenter project I've done.
Powerplant's at datacenters often use drychem systems instead or foam releasing systems.
That's not always a thing and doesn't help with fires in the rest of the complex. Backup generators, fuel depots, power distribution, large-scale cooling infrastructure all pose fire hazards even when facilities are built with actual regulation. The current AI buildout in the US happens pretty much without any meaningful regulation and gives the impression they'd build unshielded nuclear reactors serviced by children while running if they felt they could get it running cheaper and faster.
> Opposition to data centers is like 90% hatred of Silicon Valley and its obstinate elites. One they amplified by trying to build these things sneakily.
This is utter nonsense, and reeks of a desperate and deep denial. I mean, you might as well rationalize your denial by claiming they don't like you because they don't like you.
Have you tried actually listening to what critics have to say? Nimby doesn't just pop up out of nowhere at ever spot because someone failed to do PR.
> Have you tried actually listening to what critics have to say? Nimby
I wouldn’t dismiss the critics as NIMBY. They are closer to fuck it, nowhere.
Some opposition is just anti-AI or misinformed about e.g. water use. But most of them have legitimate complaints about, on one hand, noise and power prices and, on the other hand, the shocking lack of contribution to local employment or taxes from these structures minting—very visibly, they’re not subtle on social media—billionaires at an unprecedented pace.
> because someone failed to do PR
They should have engaged with the whole community and made the commitments needed to ensure they weren’t jacking up residential power prices and e.g. committing to funding the municipal budget in ways locals would champion.
I don’t understand why entry of a large consumer of electric in to an area should result in higher residential electricity prices.
The new large-consumer can pay for the electricity it uses all by itself. If new generating capacity is required to meet the new, higher, demand, that should be factored in to the supply contract with the new large-consumer.
What is the rationale for jacking up residential retail prices? And what is it supposed to achieve?
If the argument is the whole sale price increases because there’s more demand, none of that new demand is coming from the pre-existing residential users, what does it have to do with them?
> If new generating capacity is required to meet the new, higher, demand, that should be factored in to the supply contract with the new large-consumer
My understanding is this wasn't done.
> What is the rationale for jacking up residential retail prices?
The most understandable explanations I've heard had to do with grid upgrades. Crap the utility owns and controls and has to change to accomodate the new power flows.
More often, however, it looks like local and state government engaging in industrial policy by trying to get residents to subsidise the new datacenters.
Upgrades also take time and may not be a good idea seeing how the current AI buildout is financed. There is a certain amount of capacity right now and AI behemoths have the financial muscle to take what's there. If that leaves too little for everyone else, prices will reflect that as soon as existing contracts allow.
I’ve heard there’s misinformation regarding water use. As far as I’ve heard, these data centers use quite a bit. I imagine they just use them for cooking and spit it back out warmer (though if we’re going to do that, I’d prefer we do it cooling a nuclear power plant, personally…)
Is this not accurate? Are there any good resources that aren’t like… these companies being super honest about how super good they are?
Construction Physics [1] and Hank Green [2] have done good takes. Broadly speaking, most general-use datacenters use evaporative cooling (i.e. open loop) because it's cheap [3]. AI datacenters reject so much heat, however, that even absent public pressure they pretty much need closed loops [4].
The misconception is that while both of these processes use a lot of water, we both have and use tremendously more. That makes their levels tiny compared with plenty of uncontroversial, thoughtless uses. And the increasing efficiency makes this a choke point across medium-term scaling. (And replacing an AI datacenter with a conventional one counterproductive from a water-conservation perspective.)
Power—and with it, noise, air pollution and grid-upgrade costs—is the real and present problem.
The reality is that there are two approaches for water: open loop and closed loop, and you can guess correctly at what they mean. The problem is that's subtle and we can't have subtle conversations about things.
The problem is not that subtle, the only somewhat subtle point is that people don't differentiate between the datacenters of yore that were built sort of well and sustainably and the ones built in the current mad gold rush AI construction frenzy. The former aren't very problematic, often pretty good neighbors even, while the latter are nightmare fuel and will purposely not be built to be quiet and safe and stay within available energy capacity and not deplete struggling water tables, and they have the financial and political clout behind them to get away with pretty much anything. Small wonder no one wants them anywhere near them. Traditional DCs might be able to not be a nuisance despite using evaporative cooling, but no one would build that anywhere where there is a shortage of water, and the amount of heat to be dissipated by such a DC is way smaller.
But giving up a local water table might even be seen as an acceptable sacrifice, it's done for agriculture all the time, if only it wasn't done by and for the immense profit of corporations that aim to destroy all entry level jobs
and profess to be scared by what havoc their own creation will wreak, all while creating an unprecedented financing bubble.
Like, that's not very subtle at all. You can pick out some detail and try to make it about subtleties, but the larger context is very much not about subtleties.
>the only somewhat subtle point is that people don't differentiate between the datacenters of yore that were built sort of well and sustainably and the ones built in the current mad gold rush AI construction frenzy
So you're saying that all the rules and laws and best practices that HN has cheered for over the years that apply to new construction and new industry and claim to be about sustainability and environmental protections (among other things) didn't actually do what they were sold as being able to do?
Either the new ones are cleaner (I suspect they are but the difference is probably mostly on paper), or we all got lied to.
I haven't heard of many open-loop cooling systems like that in HPC datacenters. The majority I've seen use closed loop, where the water never leaves the system. Similar to most watercooled gaming PC rigs.
Nah, I back the Abundance movement and have voted for building new housing complexes in my town against groups who claimed it would ruin the character of our town to have a 60 acre plot near an old industrial center turned into neighborhoods.
Even so, given no other data I would default to no for any data center. Too many of them are doing the “move fast and break things” model while not paying for any of their externalities. If they really want to build in my location they can pay for the extra costs up front instead of claiming that somehow miraculously the economics will work out and they will definitely employ the same amount of people to maintain the building as they did to make the building.
Because people aren't motivated by facts. They make up lies to justify their opposition: "water usage". But when you disprove that superstition with facts and logic, their opinion on water usage doesn't change. Its like as if you had said nothing at all.
Raise property taxes in the burbs significantly so they can pay for their own infrastructure instead of relying on being subsidized by the poorest people in the area.
Yeah if you look at it purely from a municipal profit and loss point of view, datacenters are clearly great in that they ideally would pay higher taxes and have less drain than something like a shopping mall.
But in this case you also need to consider other points of view to decide what the right answer is, such as the extreme resource extraction datacenters have, pollution issues, and the opportunity cost of using limited land for other purposes.
If you're looking purely from a short-term profit and loss analysis and ignore all other factors, datacenters is the right answer by default but you arrive it at it using the wrong method.
You’ve forgotten sales taxes and income taxes generated by shopping malls, which is a pretty big oversight. The square meters per employee is much lower for retail than for a data center. They’ve gotten quite good at keeping staffing low.
Office buildings take land and resources and generate almost nothing for the local community, and are islands in vast seas of asphalt and traffic attractors to boot. But nobody moralizes office buildings, just data centers, despite the fact that data centers are usually objectively better neighbors.
This was a very significant talking point through most of covid, actually. Granted, there is an orders-of-magnitude disparity between office buildings and data centers regarding both their resource consumption and their local financial return through employment. Its no surprise that the attitude towards data centers is much more negative.
Not really, no? I'm not saying office buildings are "just as bad" as data centers; I'm saying they're substantially worse. It is much more unpleasant to have an office building next to your home than it is to have one of the CyrusOne Aurora data centers. The data center is --- seriously go look at them --- basically like having a park next door.
Office buildings employ considerably more people, support nearby restaurants, consume less power and water, produce virtually no infrasound, and are more attractive than data centers. I'd rather have an office building next to my house than a data center, especially if they would not be providing me colocation services at a steep discount.
Office buildings are not more attractive than data centers; that's my whole argument. In format, massing, and design, data centers essentially are office buildings, just with drastically less parking. Asphalt is the biggest eyesore in commercial real estate.
(Data centers consume immaterial amounts of water.)
Drastically less parking, so way less jobs in the local area right?
Like maybe the municipality doesn't care because a lot of jobs are filled by people coming in from an hour away, but surely the inverse of "I am sad because <some job source in my area> closed" is "I am happy because <some job source in my area> opened", right?
I get what you're saying about the buildings themselves. I just think that, in the abstract and all else being equal (it isn't of course!) "big box for 50 jobs" and "big box for 200 jobs", one seems nicer. Maybe I'm overestimating the job differential. And this is all a bit of a false dichotomy anyways
First, no, all these buildings have wildly overprovisioned parking. Most of the lots are empty. I'd encourage people wondering about this to just go drive out to an office complex in a suburb near them and look. Second, no, these jobs are not generally filled by neighbors of the office complexes, or even people from the muni where the office complex is situated.
The big box for 50 jobs is also a nicer building to have near you, at least when we're talking about the data centers you find in the western suburbs of Chicago.
Is this not the same for the data centers? Have data centers solved parking provisioning in a way office parks haven't?
> Second, no, these jobs are not generally filled by neighbors of the office complexes, or even people from the muni where the office complex is situated.
I mean again... are the data center jobs like ... guaranteed to be people in the local area? Isn't the nature of the job provisioning going to be the same with both buildings, roughly? What is special about the data center jobs to not have parking/traffic/etc?
I saw the pics of the data center you're talking about, with nice trees. I get what you're saying, it looks nice!
But it's not like if I made a nice park buffer around my office park that you would be like "oh they're the same" now right?
Is your argument "if you have a data center and an office park of the same size, the data center will generate less traffic and be less of an eyesore"? Then that means less jobs right? Why is the "cost to municipality per job" lower with the data center? Or maybe your argument is simply based on value to the specific municipality, and that the jobs are not worth much to your eyes in any case/in fact are negative, and so it's about (glibly) damage reduction?
(in other words: if you're going to a council meeting to fight a data center being built, you probably should be going to the council meeting to fight the office building being built...?)
(Serious question, I want to understand the dynamics here better)
No, it's not the same! At least not at the data centers I've seen. Go look at the Aurora CH1 CyrusOne DC and then beam yourself down 88 a bit to the nearest office building and compare. Data centers have parking, but not a lot of it. CHI1 is ringed by a single two-sided line of spots. Every big box retailer is surrounded by an ocean of asphalt larger than CHI1 itself.
It amuses me that you have consistently refused to draw the connection to this 'ocean' of asphalt and the purpose of the building it was designed for and are basically being obtuse.
But you know what is a better neighbour than a data centre or an office or a strip mall? A nice public park and local shops that support the community without you having to drive your car a distance that in any other country would be easily walkable.
Why would anybody want a monolithic megastructure sat next to their house that offers nothing good to the community itself? There is no upside for the locals. No improvement to quality of life.
You know what's even better than a nice public park?
A hospital and a university.
And if they don't generate taxes, they should fund a welfare program too.
That is what the discussion sounds like to me.
We turned from discussing whether data centers cause concrete issues to the point where authorities should intervene, to some sort of wish list for how data centers that are constructed far away from residents should provide great public benefits.
The data center might be more attractive but the office building provides more jobs and resources to the local community. You skipped that point in your reply. Even if most of the people working in the office building aren't from the area, those people are getting food and other services from local businesses. All that money goes to the local community and supports local businesses.
You'd take a featureless, opaque warehouse over a structure with windows and architecture that was intended to stand out? The asphalt doesn't have to be an eyesore if planned well, either. Copious vegetation woven throughout, painted teal or with murals instead of black, a policy of "after the following hours, locals are allowed to play roller-hockey and tennis here.", host regular farmer's markets, etc.
That’s wild, where do all the people from Aurora work? The data centers? Also what does the specific building or set of buildings you hate in one town have to do with the concept of office buildings
I've lost track of what your argument is. Mine is that office buildings are not generally filled with workers from the immediate municipality in which they're situated.
We're talking about office buildings because office buildings and data centers are both large-format commercial real estate developments.
> That’s wild, where do all the people from Aurora work?
Last I checked, Aurora had a large blue-collar Latino population, many of whom are successful tradespeople.
These people commute to construction sites elsewhere during the day, and perhaps your own neighborhood, where HN posters snitch to the town if they use a leaf blower within 1500' of their home office.
I don't know what you think this link shows, but you can just jump to CHI1 on Google Maps (or, you know, drive there) to see what that area looks like.
I do not believe this to be true! I have the impression you keep up with zoning discourse etc in your area based on your posting history... you surely have heard of people complaining about things like traffic concerns and general "ugliness" complaints around building out office parks, right?
Data centers might be making the nat'l news at the moment, but it feels hard to consider any subcategory of project development that doesn't have a bunch of whiners associated to it.
I'm a housing activist (we're fourth and inches from getting Oak Park to be the first municipality in Illinois to outlaw single-family exclusive zoning) and from grim experience I can tell you that my neighbors would far rather have an office building go up than an apartment building. It has specifically come up.
> Office buildings take land and resources and generate almost nothing for the local community, and are islands in vast seas of asphalt and traffic attractors to boot.
Manhattan (or any downtown business district) has lots of office towers with little asphalt, and many (small) businesses that make their living from office workers. Perhaps you are biased towards car-centric design thinking?
Toronto (for one) has condo towers next to (or not far from) business towers so many folks can walk, cycle, or take transit as their commute. And this is not a new concept:
> Office buildings take land and resources and generate almost nothing for the local community
This reads like the prelude to Month Python's "what have the Romans ever done for us" skit.
I recommend you catch up on the post-covid "return to office" conversation, particularly the arguments from municipalities and local governments. Even though I'm a diehard WFH guy, you clearly missed all the arguments on how forcing people to commute to the office benefits the economic activity in urban neighborhoods.
Now, if your employer opted to build offices in the middle of nowhere... That's another issue.
"Local businesses" are net money extractors. Local residents spend their money, a portion of which goes up the supply chain and out of the community.
Datacenters are net money injectors. Even if they provide relatively few jobs, or get tax breaks, they're still largely injecting capital that originates from outside the community.
>...data centers are better neighbors than almost any other commercial real estate
The biggest reason I wouldn't want a data center next door as opposed to an office building is the noise and pollution from backup generators (that may run more than expected if the datacenter has load shedding agreements with the utility)
The sea of parking lots doesn't seem any worse than the huge featureless datacenter building.
I would have the same concerns but note that backup generators aren’t just run when needed but also have a weekly/monthly quota of hours they need to run for both testing and keeping them in shape.
Approximately 80% of the homes in my neighborhood have standby generators that all need to exercise weekly. You think it would be a nightmare but you almost don't even notice when Wednesday rolls around unless you are standing outside at just the right time.
Sewage lift stations and cell towers have generators that need to exercise as well and I've never once heard a complaint about this situation. Your local municipality probably has something like 10-40% of its average load available in automatic standby generation capacity.
These AI data centers shouldn't need backup power at all. The workload does not justify that kind of reliability. If you've got tens or hundreds of these facilities, you don't need every single one of them to serve as a resilient durable hub. You just need a few of those. An inference farm going down and having requests automatically retried elsewhere is not exactly rocket science.
Strip malls at least offer something to the local residents. I recall from my time in Chicago that some of the best immigrant-family-run restaurants were in strip malls on the northwest side.
Warehouses, though, are worse than data centers in that they bring a lot of traffic to the area. (I currently live in Riverside Country, CA which may be the warehouse capital of the country).
But if we're being honest, most people see warehouses as a way to get cheap stuff to them, which they like and data centers as a way to make them unemployable, which they don't like so much.
AI company leadership are telling everyone, and that's paraphrasing only lightly, that there's a chance AI will destroy humanity, and will probably cause mass unemployment, so we need to put datacenters everywhere.
They are also universally saying that AI's benefits far outweigh the challenges and will produce extraordinary improvements in prosperity, health, and human capability.
> will produce extraordinary improvements in prosperity
Who says that? Who is doing _anything_ for AI to result in (even non-extraordinary) improvements in prosperity for the average citizen? I'm hearing a lot about AI reducing need for human labour, but literally nothing about doing anything to transition into a post-labour society. Absolute madness.
So far it's all very standard (if scaled up) wealth concentration, and not even a pretense of it trickling down.
>But data centers by themselves intrinsically seem like very good neighbors, much better than warehouses and strip malls.
It totally depends on the developer. Ask someone in Memphis about data centers and they are unlikely to mention energy usage/costs or water usage and will be hyper focused on being a bad neighbor like air pollution and noise
For sure DCs can, and probably mostly are built in a way that if there's a good thicket of trees between you and the DC you wouldn't know it's there, but there are plenty of bad actors building things that ruin the area and that's what gets all the press
> The data centers in Aurora are surrounded with prairie green space
I'm skeptical that they're actually surrounded by American prairie. What are the addresses of some of the data centers you're talking about?
Edit: I just see traditional American landscaping surrounding the CyrusOne data center in Aurora. I don't really see any native forbs or native grasses and it has a bunch of trees... where's the prairie?
> The funny thing about data center concerns in Chicagoland, where I'm hearing it, is that data centers are better neighbors than almost any other commercial real estate.
Data centres are better neighbours than a steelworks, or a pet food factory, or a fish processing plant, or an oil refinery, or a big distribution centre. By the standards of industrial zones they're low noise, low smell, low traffic, low injury, low pollution, low disaster risk. Low jobs and low tax too unfortunately, but you can't have everything.
The thing is, data centres are often built on open fields. So from the perspective of locals it's not data centre vs dark satanic mills, it's data centre vs open countryside.
> The data centers in Aurora are surrounded with prairie green space; everything else, including the nicest office buildings, are a sea of eyesore asphalt parking lots.
The concern about datacenters isn’t about visual aesthetics or traffic (the relative absence of workers—which are the traffic and parking lot drivers—for the footprint is actually one of the concerns, though not na environmental one) so this seems to be less of a contradiction than it is implied to be.
A data center can be a much quieter neighbor than a warehouse or shopping center. The catch is that most of its impact is pushed into less visible systems
Where in Aurora? I’m at the CME DC daily, and I am not aware of any prairie green spaces that aren’t being actively developed. In fact the area between Bilter and Butterfield Rds are so active it’s changing my commute time to the DC — making my 10 minute ride often more than 30. So as someone who visits Aurora data centers daily I’m not seeing it.
The point you’re making seems to be of “lesser of two evils” variety, aka DCs are less evil than strip malls. Debatable, but the lesser of two evils is still fucking evil.
I don't understand his point either. I used to work off of Diehl Road not too far from the CyrusOne DC[1] and it is anything but a prairie. It has greenery, but it's just the usual fake landscaping that all the industrial parks in Aurora, Bolingbrook, Warrenville, etc. do.
And there's counterexamples to the desolate office space... the Ace Hardware corporate office[2] I think is remarkably pleasant.
Plot a line from CHI1 to all of the nearest houses and get back to me with the the prices. I did one ray from CHI1 to Molitor Road and the first house I hit was over a half mil. Tell me more about the poor people these data centers inconvenience!
> The funny thing about data center concerns in Chicagoland, where I'm hearing it, is that data centers are better neighbors than almost any other commercial real estate.
This baseless assertion contrasts with any report or reasoning. They are noise generators operating 24/7 that eradicate social activity and purpose to such a degree that makes high-security prisons feel like community centers.
Please in your own words state what you feel is the primary advantage of having a data center in your neighborhood.
> The data centers in Aurora are surrounded with prairie green space; everything else, including the nicest office buildings, are a sea of eyesore asphalt parking lots.
You don't need to build a data center to plant some trees. There are other ways to do that, you know?
> But data centers by themselves intrinsically seem like very good neighbors, much better than warehouses and strip malls
You got to be joking. Datacenters are massive warehouses whose main output is noise 24/7.
Strip malls at least serve a purpose in the community, such as fulfilljng needs of those who go there to shop, and simplify the job if small business owners to setup and maintain a shop.
No. You can drive right into the parking lot of a CyrusOne data center in Aurora and listen. You're not going to hear anything.
Nobody is disputing that Elon Musk or whoever has built data centers powered by jet turbines or whatever, or that those are good. I'm being specific about what I'm talking about, and you're responding with a platitude.
Buddy I am writing this from that parking lot and I can hear quite a bit. I call bullshit. And suggest a hearing test, for real if you’re a DCO person get it checked on the regular.
Respectfully, I've done this recently and I don't believe you, but I'd be happy to place a bet with you and then drive out there and take decibel readings (I can go grab some hardwood at Owl to justify the trip).
This is plainly wrong considering many people's complaints about the ridiculous amount of noise these things spew out.
The whiring of gas turbines and cooling systems of complete, fully operational, data centres has nothing to do with construction.
I think trains are ok with trees. I checked satellite view - there are a lot of rail tracks in Chicago going directly along tree lines. The rail yards indeed look like tracks on concrete with nothing else, but when I look at German ones - there are trees and bushes.
Symmetric cryptography isn't based on complicated math the way asymmetric cryptography it is. The right way to think about symmetric cryptography is that the core hard problem is simply making PLAINTEXT XOR KEY work, efficiently, with a key that repeats.
Isn’t another aspect of it that it’s sufficiently random / unrecognizable, for example?
I’m very much aware of the differences between symmetric and asymmetric encryption, and realize that symmetric encryption is much simpler, but I figure that if there are weaknesses to be found in algorithms such as md5, then surely there are also potential weaknesses in symmetric encryption algorithms?
Now I’m not saying that this would be the case for battle tested algorithms like AES. But is there any particular reason why this whole category could not possibly have weaknesses?
You can certainly invent a weak block cipher, and an LLM would probably do a decent job spotting e.g. something that could be productively attacked with a SAT solver.
There's a _lot_ hiding in that, all of the interesting stuff for security and potential breaks. So...yeah it is based on complicated math, it's just in that bit instead of the xor.
Even the xor is a bit of a fudge, but probably close enough.
There's a lot of basic computer science hiding in it that's been remarkably stable for generations of computer scientists, which is not something you can say about asymmetric cryptography.
Is there? Like...kind of, but on the face of it I'd say about the same amount in both. If you look back at DES there's a _lot_ in common with modern ciphers, but like, RSA is still in use and that's old as shit.
I think you're right if your point is that we're more likely to see big breaks in asymmetric crypto, but it's kind of based on vibes to me, it's not really clear that it's provable in any way with anything like our current understanding.
AES doesn't reduce to a fundamental mathematical problem we're uncertain about, in the same way as discrete logs, factoring, the elliptic curve discrete log, or shortest vectors. It's a simpler idea, mathematically: rigorously understood linear operations to propagate key-driven changes quickly, disrupted by nonlinear operations to keep the cipher from being solvable with algebra, driven by a key schedule, and iterated enough times to destroy the signal that differential cryptography (and its analogs) would use to mount attacks.
It's just radically different levels of exposure to mathematical theory.
I'm fond of pointing out that JP Aumasson, who is (unlike me) an academic cryptographer of some repute, believes SHA2 will never be broken.
For any kind of symmetric encryption method, if you are in the context of a "known-plaintext attack", i.e. where besides the cipher that was used you know both the non-encrypted text and its corresponding encrypted text, if you know a segment of text that is long enough, i.e. longer than the number of bits corresponding to the secret key and of some ancillary information that might be needed, like an initialization vector, the problem of breaking the cipher is known to have a unique solution, because it is equivalent to the solution of a system of equations that has enough equations to determine the unknowns.
However, for a well-designed cipher that system of equations is non-linear and exceedingly complex and it cannot be formulated in any mathematically simple form, therefore no methods of solution are known.
The more rounds are performed inside the encryption algorithm, the more complex that system of equations becomes, and the more unlikely is to find any method to solve it.
Even for the much simpler systems of non-linear equations that occur in physics, when they cannot be linearized even in the neighborhood of each point, then there typically are no better methods of solution than searching randomly through the solution space, which in cryptography corresponds to brute-force search, which is infeasible for big enough secret keys.
>if you are in the context of a "known-plaintext attack", i.e. where besides the cipher that was used you know both the non-encrypted text and its corresponding encrypted text
I know with some pre-modern ciphers like Caesar that was an issue, and it famously came up due to implementation/operator issues with systems like Enigma. But is that actually still a relevant attack context at all with modern symmetric encryption? Outside of academic interest I'm struggling to think of any time where you would ever be in that situation. I thought input got scrambled pretty well so you can't really trivially predict things like text word count (putting aside that these days the vast super majority of bits aren't text). When used to protect your own data (ie, storage encryption etc) then the "end points" are you and future-you and by definition fully trusted, including operation of the encryption system bits. If someone hostile roots it and has access the device computation they'll also have access to device data. When symmetric encryption is used between end points controlled by two different parties keys forward secrecy should be pretty standard these days right? So attackers still would have to root one trusted side or the other and at that point it's game over regardless.
On the face of it seems like it'd have to be a pretty strange situation nowadays where the attacker somehow gets to reuse the same encryption key as the trusted parties without knowing it or being able to access the secrets they're trying to keep without touching the encryption at all? DRM sorts of use-cases maybe, one of the few situations where hostile attacker and trusted user are supposed to be the same person which has always been a tension with it.
reply