Most regular people have not been sold on the idea that more AI will be good for them or their families or communities personally. Instead all they hear is AI taking away jobs, and AI preventing high schoolers and college kids from learning how to read a book and write an essay. Most people think a world with more AI seems bleaker.
So people wonder why communities don't want this hulking, loud, energy-expensive data center whose main benefit is to hasten the arrival of this world they don't want?
If you need your job to feed your family then what Sam or Dario is saying is quite literally a threat to kill your children.
Of course they’re full of shit, it’s more like a very useful business automation project but if you say stuff like that you can’t be surprised when people hear it and form highly rational opinions about it.
Isn't AI the only way we get to that? Who is gonna do actual important work otherwise?
Instead the evidence of our own eyes tells us that AI seems to be worsening the K-shaped economy: a handful of Elon Musks will become trillionaires, a slightly larger number of engineers will become billionaires, and the rest of us will have fewer jobs and fewer prospects of upward mobility in a world where AI can do most entry-level work.
You don't get to pick and choose what productive technologies to adopt. You either adopt them, or you'll be paying the countries that do and you'll still lose your domestic industry to competition.
It could make economic sense to be the world leaders in the next drug that turns everyone into mindless worker drones ("if we're not making the factories that make the drugs the Chinese will!") but people still won't be happy about what it means for their futures.
Until AI proves a systemic risk to humanity there would be no similar consensus on shunning its use, and the economic spoils go to the countries that adopt it.
* use a lot of water to build one
* the power is a problem
* the noise is an issue
* the power is coming from non-green sources
* the government could “easily” fix all this with regulations
Also can you tell us which chemicals need to be added to the outer and inner loops of the systems, and do the inner loop has the same losses or it is different.
If there's water loss into the ground, and if the water is treated with chemicals, is it required or is it in fact treated to remove these chemicals so the used water can be reused for livestock, agriculture, or non-consumable use in the cities like for watering greenery or other processes which needs water but can accept non-potable variety.
I also heard that there's a closed loop version for outer loop. Is it used widely in US, if yes, what percentage of the large datacenters use it? Because as I heard it, it's more expensive, but doesn't lose water as much.
Adding some references if possible is greatly appreciated so we can dive deeper, as the community.
Same is also valid for power usage metrics. I'm especially interested in how a datacenter full of 800VDC systems (aka Megawatt Rack) can be fed without any difficulty with our current power generation and transfer infra.
If it doesn't, we can all stand around and make fun of the people who paid terabucks for the data centers, and enjoy a bountiful harvest of surplus hardware for pennies on the dollar.
OTOH, considering the proliferation of the EVs, greenhouse gas emissions from cars are going down, alongside brake-pad related emissions since generative braking covers 90%+ of the daily braking needs.
OTOH, many datacenters are installing their own greenhouse emitting power gensets. So, with a bitter contrast, as AI is replacing jobs, it's also filling back the carbon footprint vacated by cars and more efficient modes of transport at the same time.
These GPUs won't be scrapped soon, because besides AI, they are useful for accelerating not-so-precise floating point calculations en-masse. So, no, it won't work as you imagine.
And here we are, talking about the US. Great, that's settled.
OTOH, considering the proliferation of the EVs, greenhouse gas emissions from cars are going down
In the US, most electricity comes from carbon-based sources. EVs help, certainly, but it would still be best from an ecological standpoint if people didn't have to use them every day to get to work.
OTOH, many datacenters are installing their own greenhouse emitting power gensets.
Exactly, which is why we should work for grid renewal and expansion at the same time we build these things out. If the hyperscalers insist on building their own power plants, those power plants should be required to be (a) carbon-neutral; and (b) grid-tied.
These GPUs won't be scrapped soon, because besides AI, they are useful for accelerating not-so-precise floating point calculations en-masse.
Really. What are some other applications for not-so-precise floating point calculations at the exaflops level?
Proliferation of law skirting AI datacenters is a worldwide problem. Many companies are extending this to Europe as well, so AI datacenters is not a U.S. problem only.
From what I see in Electricity Maps[0], it's not clear cut. Some of the states have well over 50% of their generation is from carbon-free sources. As a general trend, west is greener than eastern US. So, no, EVs will have an impact there. However, you need electric semis a lot as a country since rail is not used a lot in the US as well.
> Really. What are some other applications for not-so-precise floating point calculations at the exaflops level?
In short, HPC. As a more detailed, by short list:
- Weather and climate modeling.
- Drug discovery.
- Computational Fluid Dynamics.
- Material science, behavioral and process analysis.
- All kinds of large scale simulation which can be modeled with matrices.
- Any kind of problem requiring linear algebra.
While I generally work at full 64bit precision, from what I see, even though these modern GPUs do only support up to FP16 (or FP32 in some cases), they can be used for higher resolution simulations as well. We support some astronomy and climate researchers which use classical algorithms and in-house ML models (not LLMs and such) to accelerate their simulations and predictions on multiple GPUs.I manage HPC clusters, and sometimes develop code on them, so yeah.
[0]: https://app.electricitymaps.com/map/zone/US-NW-PACE/live/fif...
(IMHO, the fact that AI models tolerate absurdly-low precision means that we're doing something wrong and wasteful in our entire computational approach to the field, but that's neither here nor there.)
Except when some rich clown from another country comes into a drought-stressed desert area where residents have already been under increasingly draconic water usage restrictions for goin' on a decade now, wantin' to build a datacenter right on a lake that's already under severe stress with record low water levels, tellin' everyone it won't have any effect on anything and it's all just rainbows and sunshine and nothin' but benefits for everyone all around (lies). But yeah, no worries. We can surely trust the billionaire wouldn't lie to us to enrich himself at our expense.
> In any case, even the noise worry is overstated: many viral videos of supposed AI data centre noise are really Bitcoin mines, not state-of-the-art hyperscaler facilities.
Are there really major differences between a hyperscaler, neocloud, and crypto mining data center in terms of noise pollution or other negative effects?
You even note the problems, but then wave them away saying that "the government could fix them."
So have the government fix those problematic regulations first, and then the AI centres can re-apply for permission.
But what people are most upset about are water and power, which are the two least problematic things about datacenters.
(This is part of why I no longer help build them, because when I asked about reducing noise and using green power I was laughed at).
I think this is the core of the 'moral panic'. it isn't anti-AI, it's the fact that many AI datacenter builders have been somewhere between surreptitious and intentionally illegal in their avoidance public knowledge of what they are doing, most likely to avoid exactly the regulation you're recommending. As a result, there's now categorical public distrust of both the industry and the local government that has shirked their obligation to the public welfare.
It is reasonable to conclude that you don't act as secretively as these builders have unless you have something to hide.
Like.. you came in invalidating the people, then, when called out, you say what is effectively "well their criticism is actually right".
And.. That's that. I guess. Are you sure that you're taking part in the conversation and not just stringing words together based on probabilities?
It seems with all the other battles to fight, that moving it 2km down the road and surrounding it with trees and parkland would knock the noise argument out of the window. At least in the US and Australia that really do have vast amounts of open space.
They are powering it with turbines right, so if its next to your house it sounds like you're permanently in on the airport tarmac. Not great. I saw a video (not on the internet, but from them). It was pretty annoying. But it makes sense, there is no way a power delivery system in a developed country adapts so fast to such intense energy needs. Here in india, many 250Mw datacenters are put in places where a proper grid connection is itself coming for the first time, so the same problem is not there. And frankly there are way more noisy industries in poorly zoned mixed-industrial-residential areas here for generator noise to even matter.
Soooo
You're building them, and you're agreeing that the biggest points of criticism are valid.
But.. some handwaving that the government should do something about it and also others are too doing bad things.
And your conscience is clear, huh?
Talking to just a few people or one person at a time can help a lot, also.
But I hear you. It sucks trying to be helpful and then getting it shoved in your face.
But that's more expensive to operate.
So most of them don't do it.
Show me actual statistics that a supermajority of new and proposed datacenters either are currently using (for new), or plan to use, with clear financial projections for how (for proposed), closed-loop cooling systems, and I'll consider conceding this point. Until and unless someone can do that, you can point to the option of such systems until you're blue in the face, but it's meaningless.
Even if you can get legislators to pass regulation, enforcing it is a constant battle, and odds are they'll just pay a fine and keep on violating the regulations.
It's a much easier fight to keep a DC out of your community to begin with.
Isn't that just the definition of left or right leaning? People who don't listen.
The water is mostly part of a closed system. Sure, you need a lot to build one, but once it's running it's pretty water efficient. The lack of green power is a problem, but as the article points out, you now have a large customer subsidizing the base cost of delivery, since they are only one delivery point, and they are financing new sources of power, which if properly regulated would be green power.
The noise is an issue, I will admit that. As is the fact that a lot of the power is coming from non-green sources. But the government could easily fix these problems with regulations that apply broadly, not just to datacenters.