And this is exactly how things in large corporations don't happen. Things in large corporations who pay so well that nobody makes it through the recruitment filters and competition who's more than 2% off from being a pure assessment metrics maximiser. In my mind, chasing an idea like this just does not match the type.
Why it isn't likely to have happened before: trying is only easy for a developer who is not only very familiar with the API for reading those motion sensors but also with the API for system wide overlay. This a strict no for your run of the mill to-do list app. Combine that with the requirement of being affected themselves (or someone close), having the idea, having the time to try and no competing side quests that take up all attention. And the requirement of when the experiment was successful, having the kind of access that actually serves as a conduit for ideas not only on paper to someone in the position to make it a product. That's a lot of stars that need to align. Personally I wouldn't even be surprised if there was an LLM angle: would still need someone in the know to try it, but chances are that knowledge combined with a bit of prompting could be a path to PoC implementation that could clear effort thresholds that would stand in the way of conventional coding.
That has nothing to do with the feature being “ideated, then prototyped, then validated”, which is the crux of the conversation. It having taken so long is but a footnote.
> Personally I wouldn't even be surprised if there was an LLM angle
I would bet money there wasn’t. The feature was introduced in early 2024, meaning it has to be thought about, developed, and tested before that.
Nice to see some numbers. So for the peak load situations, a Dash 8-100 would not require a battery bigger than that a short range BEV ("city", though in reality the short range BEV use case is more for the rural equivalent of stuff that would be walkable in a city setting). And that's even before considering the energy contributed by the fossil fuel engine.
"13000 * 9.81 * 1000 = 127.5 MJ, to reach your altitude"
Presumably quite a bit of that would be harvested back during the descent that follows. The conventional engine would still need some excess power (relative to cruise load) to fill the gap left by drag and imperfect circle efficiency of the electric motor/generator, but mass x altitude is stored energy, not lost. (I'm still talking about the "what if we need a second abort" of the root post)
Easy: you don't try the second landing approach before the battery is sufficiently recharged to contain enough energy for the second abort. Chances are this does not take any longer than going through the pattern anyways.
The saving is not just the dead weight of the bigger engine you'd need to do take-off, climb and abort without electric assist, it's also the fuel saved during cruise from running an engine that is completely designed for efficiency at cruise load instead of for some compromise between cruise efficiency and sufficient peak power for start and abort.
Not when they are properly used as a fossil fuel source. As a fossil fuel source that just happened to also make itself useful in other ways on the voyage from mine to smokestack. As long as we still burn coal for electricity and/or heat it's just stupid to bury plastics (or, worse, let it slowly decay unburied)
It wouldn’t be so bad if you buried it and it stayed buried it, but whatever disposal technology you use it doesn’t work if the plastic is never properly disposed of. Even if people are really careful, for instance, some single use plastics are going to blow away in the wind. Something shoddily made in the bottom of a pond is going to come apart and shed plastics, etc.
Misleading title: they are using eminent domain to seize land for power lines to data centers. (emphasis mine)
This is still quite debatable, but so much less so than the implied accusation that they would somehow sell out their privileges as subcontractors for the data centers themselves.
To be even more nuanced, the land is not typically taken entirely. The more typical solution is an easement with compensation for the land owner to cover the reduction in utility.
Transmission lines are pretty harmless when considering things that could be built instead. Zero emissions, silent, no illumination, etc.
I was looking at buying a house with big transmission lines crossing over the property. It was a misty not-quite-raining day when I went to look at it, and you could hear the lines crackling and humming the whole time you were outside. I figured I’d have to give up ham radio if I moved there, ha.
Over time as some dirty coats the insulators, and on high humidity days some kind of crackling is clearly audible in a fairly wide radio close to the towers.
Also, transmission towers are usually the best lightning rods for miles in a given location, and it is fucking unsettling to live close to them during storms. Not to mention they are an eye sore.
There's a high-voltage line that crosses over a park (with soccer fields etc...) near me, and you can hear it crackling whenever you're there. I'm sure it's a noisy radio environment.
These are not silent nor zero emission. These do emit large electromagnetic fields that interfere with some electronics. There are also some studies that show higher risk of childhood leukemia and potentially negative affects to sensitive groups such as people with neurodiverse conditions. They also tend to hum and crackle. At times you will have helicopters doing line work too.
> potentially negative affects to sensitive groups such as people with neurodiverse conditions
People with neurodiverse conditions are not sensitive to power lines. Hard to believe we’re seeing this type of claim on Hacker News.
Modern high voltage DC transmission lines also don’t have 60Hz emissions because they’re DC.
Anyway, the land where these transmission lines are built usually isn’t running through high population density areas. They’re not slapping a HV transmission line right in the middle of a residential corridor if they can avoid it. It’s going through farm land and empty space if at all possible. It’s a hundred times easier to build the towers in those locations instead of near population centers.
you've never walked under a high voltage power line?
even driving a car under it you can hear the radio being weird. i assume it can make life miserable for a ham radio enthusiast.
also, walking beneath one, you can hear the crackle, which is why nobody places a power line above homes.
as for them causing leukemia i'd not consider that realistic
My home is like 1500 ft from HV lines. According to OSM, they are 115kV. Some of my neighbors are even closer. These things are totally built near homes.
Many of them probably are. Individual EMF concerns are often debunked. It is possible but unlikely that all of them are categorically complete nonsense. But let's put them aside for a moment. There are real, known consequences of living next to a data center as implemented.
Coil whine from transformers, in audible and infrasound movement of air, has real known effects, as well as suspected poorly studied effects, on the human head & ear.
The fact that we're just allowing a data center that says "I'm out of power, let me use more", to run four extraordinarily loud, dirty airplane jet turbines at the middle of their property for power generation on an improvised basis. And then when we look at their operation in aerial photos, we see 36 jet turbines.
Any and all non-closed-loop cooling systems are an issue, especially when they pull from potable water or operate in dryland droughts. The ratio of wattage consumed vs water consumed should be instructive here.
The amount of money that is going into these things is obscene, and when they choose to involve local officials it is essentially buying out your local-level democracy wholesale. Total lack of regulation. Town council meeting videos where all of your supposed representatives are under NDA; Extremely blatant corruption that the state doesn't want to touch. Races to the bottom.
These things are designed to run basically as lights-off factories, with little to no local labor involved. While there may be some initial investment in facilities, that will dwindle. When they are simultaneously not paying taxes via assorted individual arrangements, and raising electrical rates, there's very little to recommend them to locals.
We have established no general rules as far as buffers. Large-scale construction in general tends to kill rural watersheds & aquifers in close proximity. The sound issue is real, the EMF issues may have some limited credibility, and well-water is often tragedy of the commons, all of which might be totally mitigated by a half mile or whatever of forest buffer mandated around a data center zoning.
I am going to say ok maybe possible but I look at how many transmission lines criss cross places like Dallas. Would there not be enough data to have already proved that thesis?
I remember the original issue with these is they were spraying some kind agent orange like chemical to keep growth down under the lines and people were getting cancer from it 50 years ago.
I've got overhead transmission lines in my back yard right now. It's really not a big deal. I've got a garden and a small shed that are directly underneath the lines. I have no perceptible loss in utility. Every now and then, someone will visit my property but I don't have to interact with them.
The nature of the easement can make a big difference. Oftentimes the power lines will be colocated with gas pipelines which tend to impose stricter requirements on land owners. It's a lot easier to do things like excavate when you can see what you aren't supposed to hit.
The buzzing is pretty obvious when it is rainy or misty. I passed a power line during my high school bike commute, it never failed to amuse me. I wouldn't call it "loud" though: noticeable, but the average traffic noise is far worse.
The leukemia is bullshit, though. The numbers are too small to make any kind of definitive statement, or to properly compensate for things like socioeconomic factors - plenty of other potential causes on the rural farms and poor urban neighborhoods where we usually build power lines, after all.
> The leukemia is bullshit, though. The numbers are too small to make any kind of definitive statement, or to properly compensate for things like socioeconomic factors - plenty of other potential causes on the rural farms and poor urban neighborhoods where we usually build power lines, after all.
This seems like the thing where people use the phrase "there is no evidence of" to imply that something has been disproven when what it actually means is that we don't know enough to say one way or the other.
Basically I just thought the leukaemia claim is quite an extraordinary claim, which requires extraordinary evidence if you’re trying to convince a sensible person of your claim.
I know a thing or two about rf, emf and electricity in general and I don’t see why or how these hv wires would cause cancer, or if they did why would it be specifically leukaemia.
I’m fully aware though that before we knew about nuclear radiation people would have had similar thoughts about some random rare earth metal causing cancer if you’re not even touching it. So point is, I’ll wait for the proof before I’m convinced, or attempt to convince others.
I’ve also anecdotally noticed that people are concerned about many things where there is like very marginal evidence that maybe this is slightly bad for you but then regularly engage in very well known risky behaviour like eating big macs or driving cars. I might be speaking from personal experience.
> I know a thing or two about rf, emf and electricity in general and I don’t see why or how these hv wires would cause cancer, or if they did why would it be specifically leukaemia.
Probably the same theory for how a microwave cooks food.
> I’ve also anecdotally noticed that people are concerned about many things where there is like very marginal evidence that maybe this is slightly bad for you but then regularly engage in very well known risky behaviour like eating big macs or driving cars.
Eating junk food has the countervailing benefit that they like the taste. There is very little personal benefit to having a high voltage line in your yard, and people respond especially poorly to things that have a personal risk with no personal benefit.
"Probably the same theory for how a microwave cooks food."
The actual change to look at isn't really the heating, but rather that RF has been known to increase glucose metabolism in affected cells. My own theory is that if that is true, you're essentially increasing cell aging and oxidative stress with the increased metabolism. Who knows if it's true. The effect seems small and the exposures need to be longterm.
Surely power lines running directly across a residential property represent a long-term exposure.
(I personally doubt there is a significant risk, but if there is, these powerline easements would be quite risky to the residents in terms of ongoing/continuous exposure.)
> what it actually means is that we don't know enough to say one way or the other.
To take another illustration: nobody can prove that there is not between the Earth and Mars a china teapot revolving in an elliptical orbit, but nobody thinks this sufficiently likely to be taken into account in practice. I think the leukemia claim just as likely.
If there was a teapot at a proposed orbit then there would be various falsifiable ways to prove it one way or the other. Once you do the experiment, you have the evidence. Before you do the experiment, you don't.
If there is a cancer risk then good evidence either way would require randomized controlled trials, which for long-term exposure are very expensive and take a long time, so then nobody does them and nobody really knows the answer. This is not the same thing as knowing it to be false.
The crackling of HV-lines was used as a sound-effect in that game. 'Realistically'. Using that buzz and crackle from reality, only slightly ampflified.
Is the nuance you're contributing here substantiated in the proposals being discussed in the article?
Transmission lines emit EMF which seem pretty harmless unless you're a pollinator or migratory bird. Maybe the home owners are just making a fuss so they can get a better payout to improve the amenities in the bunkers that comprise their subterranean community of concrete beneath the barren badlands of the over-world. At the very least, they should ensure respirators and gaussmeters for able-bodied coal miner minors.
There's a lot of people saying they aren't harmless but there's no credible evidence they aren't harmless. It's like flat earth, you know, there's no evidence the earth is flat.
Actually, strictly speaking, it's /more/ debatable. Using eminent domain to seize land for the data centre itself gets a much more generally negative reaction, while seizing it for the purpose of infrastructure to supply said datacentre - and therefore, at least arguably, alleviate strain on the infrastructure used by everyone else - has more nuance and more room for debate.
Data centers are a fixed load with on-site backup generators that can be used to shed load when the grid is stressed. Grids actually like that sort of thing. It lets them use more cheap renewables.
Dunno why you think it's obvious that I should be able to have the state seize your house for my own personal infrastructure just because I demand I chew through so much electricity that the local infrastructure can't support my demands.
Do you have the same objection if it for a housing project, built by a private corp?
That's a practice that exists everywhere. This is an article about a case of expropriation in the Netherlands. The government frequently uses expropriation ("onteigening") to forcibly remove old, small wind turbines owned by private farmers to make room for massive, industrial-scale wind farms. In Dutch, so you'll need to swith on English translation:
There have been some pretty egregious misuses of eminent domain in the past, but using it for infrastructure is pretty much the quintessential justification. That's basically the only reason to have eminent domain at all. Arguing against it when that's how it's being used is, in my opinion, essentially arguing against the very concept of eminent domain. Which one can totally choose to do! But they should be up front about it, if that's what's going on.
As for the title, I would go further than you do. This is not a misleading title. It is a lying title. I'm usually a pretty big believer in Scott Alexander's Bounded Distrust [0] concept for the media where they very rarely actually, technically lie. This title (although not the article itself), in my opinion crosses that line.
We run power lines to everything else: houses, offices, commercial buildings, factories, and so on. Any particular reasons we should have a special exception where we don't run them to data centers?
Where I live pretty much every piece of land has some sort of easement on it: for water, power, or something else. Seems weird to me if we can insist that now everything needs to be frozen and no more infrastructure upgrades.
Agreed. Since most land is already owned by private individuals do we freeze everything as is and do no further changes [1]?
The last house I lived in had an easement along the back property line. I built my fence on the easement knowing full well they would tear it up if they needed to access anything. I live in a denser town now where the power comes from the street to my house, wraps around my house, then lines to the next building over. I guess I should tell those people in the other building to figure out a different way to get power.
[1] Also feels a bit like the arguments against wind power.
But no, we don't seize rights and access to privately owned land for houses, offices, commercial buildings or factories unless there is overwhelming public benefit. I can't just say I want to build a factory and force access and utilities across my neighbours without their permission or an extraordinary process to prove that I'm going to provide massive local benefit (or national security).
If my factory wants to compete on an already contested resource (water, power), I should have to go further to show that I'm going to be a net benefit in my area.
As things stand, datacentres provide very an extraordinarily low employment density. They don't produce food. They consume an extraordinary amount of energy making it more expensive. And they money they generate concentrates wealth to billionaires. If you want your powerlines prove you're going to be beneficial. Start with a solar+batterty farm. You can surge from the network, and sell on slow days. Guarantee free power and network for locals. Integrate heat pumps to heat local homes in winter.
I've wandered off the topic of forcing permanent access to others land into broader planning, but it's all related.
>But no, we don't seize rights and access to privately owned land for houses, offices, commercial buildings or factories unless there is
Ever since Kelo v New London we do.
> overwhelming public benefit.
This is just an appeal to authority that is predicated on the broad willful ignorance that the mean reader here will have toward how this particular sausage is made.
Read the proceedings. Half the time it's bullshit like "gravel pit or office park development wants an easement through undeveloped/agricultural/industrial parcels XYZ to build a road because that means they can exit closer to the highway which saves a bunch of money. They'd pay the land owner because at the end of the day it's money either way, they just think they can convince the state to stomp the land owner cheaper than they can pay off the land owner.
A blank cheque on corpo infrastructure just means that billionaires win and everyone else loses. At that point, just paint "Peasant Slave" on your forehead if you don't have a multi-million lobbying group at the state and federal level.
"All your base are belong to us" => Billionaires of USA
When that happens rename the nation to Corporate States of America as well.
But what if I want to live in a country with a growing economy and improving infrastructure? Forbidding all development does hurt corporations and the wealthy, but just in the sense that it hurts everyone, including them.
If we do want to soak the rich, can we do it in a progressive, targeted way (such as taxes) and not in a NIMBY way where we just say no one can build or improve anything?
If they pay enough, why not? What is public use infrastructure in your opinion? Any difference than building more houses (like a dense apartment complex) and extend the grid there? That's still private corp infrastructure.
There's a difference between "The state will seize your property" and "You may sell your property to the corporation at a rate you choose, and if you don't want to then they'll have to find some other land to use". In either case, development does happen, but the primary beneficiaries of the development (the owners of the data center) would have to pay the people who's land they're using (or taking!) instead of lobbying the government to just take it.
I think that unwitting land owners should not be subsidizing the AI grifters...
Losing my home would suck, but the prospect of having a semiconductor fab in my country, in my city even, is so exciting that I would be 100% on board with it.
ALL High-tech Infra that leverages "eminent domain" SHOULD be held in the name of the nation state for public good not by the billionaire or his corpo. All these private data centers that leverage eminent domain to grab land should be property of the state and only temporarily leased to the corpo.
> Arguing against it when that's how it's being used is, in my opinion, essentially arguing against the very concept of eminent domain
You're giving a blank check to use it as long as it's infrastructure, and as any simple rule, it's prone to get abused way beyond it's primary goal. IMHO the issue is nuance and balance, we should accept that whether it's applicable is an "it depends" matter.
Keeping it too simple and black and white means companies can now strategically place data centers and get the local government to seize all the land leading to it, and vice versa, local govs can nudge companies to target specific places to seize wherever they want. Especially as the plans aren't required to come to fruition.
(Submitted title was "Power companies are using eminent domain to seize land for data centers" but we've since changed the URL - more at https://news.ycombinator.com/item?id=48975572)
It's more misleading than that. The power line that "may" "primarily"[1] serve a private data center has started construction "after nearly 20 years of planning, permitting and lawsuits"[2].
Per usual, NIMBYs continually move the goal posts and rely heavily on rhetoric to oppose a project. Data center opposition is providing them with great PR fodder, but they would have opposed these projects regardless. This is why infrastructure is so incredibly expensive in this country.
Throw around NIMBY as a pejorative until you're blue in the face. Landowning residents have a civic duty to exercise their interests to prevent the remote monied and powerful from packaging their speculative investments as valuable infrastructure when the profits are privatized and the losses are socialized. The gambling addicts aren't going to spend a single day in prison when neglected overbuilt "infrastructure" directly causes mass casualties or pervert the shape of communities spanning generations.
Are you under the impression people are not compensated by eminent domain?
In fact, a huge portion of the sentimental press stores over eminent domain are attempts by homeowners not to cancel the project, but to increase the payouts. Often the homeowners would be quite sad to see the project cancelled, they just want to milk the settlement via litigation.
I am under the impression that they are not compensated correctly when their land is taken.
I own the land that I do because I value it more than any other buyer/current bidder. (If that weren’t true, I’d sell it to the higher bidder.)
If, instead of bidding an attractive price for my land like an ordinary buyer, you win a court case or political process that says you can force me to sell my land at a “fair market value”, that previous point establishes that I’m being forced to take a loss, as I am being forced to accept a price lower than my reservation price.
At the same time, eminent domain must exist in order to let continuous projects like power lines and highways be built. I just think that 150% or 200% of what is currently called “fair market value” is a better reference price than 100% of it.
The taking of land against the owner’s will should come at a premium, but that premium should be capped rather than unbounded.
The problem with not having eminent domain is that without it, for any project of sufficient size you'll end up with ~100+ land owners who each individually have veto power over the entire project. If 99 of them freely agree to sell, but 1 is really stubborn and refuses to sell at any price, the project is dead.
There's also a prisoner's dilemma problem. Maybe it's not a big deal paying one stubborn guy 100x fair market value for his property to allow the project to continue. But if enough people think they can get 100x what their property is worth if they just hold out a bit longer, that also kills the project.
Eminent domain solves all this by making it so the project only has to to clear a few bars the government has set for it, rather than the many bars ~100 individual landowners set for it. But in the process it also erodes property rights and kills the built-in protections the free market has to ensure such transactions are fair.
A 1.5x or 2x multiplier on the fair market value would help with that a lot and help ensure eminent domain is only used in cases where it's strictly needed.
Or alternately, I sort of like the idea of letting the landowners elect someone to represent them in negotiations with the project developer, and then having a chance to vote yes/no on the final cash offer. Then the final price would still be set by a market of sorts, but there'd be no prisoner's dilemma or individual spoilers. Not 100% sure if that would work better or not though; it is a bit more complicated and there are still some issues with it.
Condemnation proceedings are the exception. Most of the time, the entity taking the land under the threat of eminent domain offers substantially above FMV, if only because condemnation proceedings cost money and, more importantly, time.
Some states do require above FMV for condemnation, e.g. 120%. After Kelo in the early 2000s, almost every state passed significant eminent domain reform.
Humans are dumb. Near my current city folks are debating a solar farm going up. Far enough out of the way that it has zero impact to anything he city, middle on nowhere farmland. The math was something silly, for farm use it would generate $15,000 or solar farm $1,000,000. I am probably embellishing the number from memory but it was staggering. Folks are still trying to fight it. It’s frustrating.
Farmers don’t like lines because you have to allow for an access road to it and it does not pay anywhere as well as say a cell tower.
I'm not a friend of nationalism, but I believe that it's a trade-off: of you want to be open to immigration, of the kind that pulls in newcomers, inviting them to become a part of the place they move to, instead of remaining outsiders, you have to give them plenty of opportunity to identify with their new home. Of course these days, we see the American flag used a lot in ways completely opposite to this, but that does not change the great progressive value national symbols could provide.
I read about this so often, yet it's a problem I just never encountered. If I need super high bandwidth like connecting a display, I pick one of those annoyingly unlflexible fat cables. Easy. They even tend to feature the TB4 flash icon. Surely I would not expect any of the light and nimble ones to do that trick. When I need strong PD, it's also either one of those or one of the far more flexible but still quite thick braided ones. Often they have some hinge gimmick connector to prevent any hope for bandwidth one might be tempted to have. All other cables, I expect nothing but legacy USB. Yeah, and some of those won't do data at all (curiosly those are never the lightest cables in my stable, the lightest ones tend to do the legacy USB "gear not, some bits will eventually get through!" just fine)
Will the workloads exist? As in not going straight to something at least as dedicated as the Strix Halo CPU/GPU combo (with air quotes around the G)? Or the Apple max? Somehow I don't picture this as an attempt to make a full AI rig that just happens to be x86 in the housekeeping parts but just something that will make CPU inference a little less bad. In that case it would at best be a hedge against some low requirements use case becoming more important than expected, yet another unused spec sheet checkbox engineering marvel otherwise.
Not long ago, China has demonstrated a supercomputer faster than any existing US supercomputer.
The Chinese supercomputer uses custom CPUs designed in China, which implement the SME matrix extension of Armv9.3-A, while the fastest US supercomputers use GPUs.
The Chinese CPUs with SME have a lower energy efficiency than the best NVIDIA datacenter GPUs, and also a little lower than the best AMD datacenter GPUs, but not only they reached a greater absolute performance, but they also reach a much higher percentage of their theoretical maximum throughput, and this is likely to happen for all problems solved on them, because writing efficient programs is easier for CPUs than it is for GPUs.
As discussed in TFA, the Arm SME and the Intel/AMD ACE are similar.
Thus server CPUs that have hundreds of cores and ISAs with matrix extensions can be very competitive in performance with GPUs, even if it is very likely that GPUs will always hold the records for the best energy efficiency.
The Chinese CPUs have fast HBM memory, with an 8 TB/s memory bandwidth per socket, equal to that of the latest AMD Instinct datacenter GPU and much greater than that of most less expensive GPUs. Very few consumer GPUs exceed a 1 TB/s memory bandwidth.
So at least where electric energy is cheap and abundant, running AI inference or training on server CPUs may be a competitive solution, especially when that computing cluster is also used for other purposes, not only for AI.
Dedicated PCIe or memory-attached inference units will continue to exist, and will continue to do the heavy lifting, but an ISA extension provides latencies any external unit would have trouble matching. You can, with some work, extract some decent throughput with CPUs alone, with a large enough CPU you can use for non-AI jobs the rest of the time. There was a nice writeup not that long ago here on HN describing the flags and the reasoning behind them to use on that specific machine.
Yeah, I guess in these days of hardware not getting completely outdated even after half a decade, it might well make sense to spend chip area on specialist units that are shut down almost the entire time.
In the old days, when it was all about squeezing maximum performance from a given chip area budget you'd prefer using those transistors for some improvement of the general purpose parts, no matter how marginal. But now it's the thermal envelope, or power consumption of you prefer looking at it from that angle, and even slightly esoteric special purpose units can start making sense.
Dark silicon is also a thermal management strategy. Areal usage can go down, spreading heat generation over a larger space and making cooling the chip easier.
Why it isn't likely to have happened before: trying is only easy for a developer who is not only very familiar with the API for reading those motion sensors but also with the API for system wide overlay. This a strict no for your run of the mill to-do list app. Combine that with the requirement of being affected themselves (or someone close), having the idea, having the time to try and no competing side quests that take up all attention. And the requirement of when the experiment was successful, having the kind of access that actually serves as a conduit for ideas not only on paper to someone in the position to make it a product. That's a lot of stars that need to align. Personally I wouldn't even be surprised if there was an LLM angle: would still need someone in the know to try it, but chances are that knowledge combined with a bit of prompting could be a path to PoC implementation that could clear effort thresholds that would stand in the way of conventional coding.
reply