Um, we solved this problem years ago, we charge industry for dirty power produced and that incentivizes the industries to install power line filters and capacitors and if they don’t we use the extra they are paying to clean it up. They might not be having to pay for dirty power since its commercial not industrial but all that requires is forcing them to be in industrial zones or changing the electrical price structure, this is a solved issue
This map shows readings from about 770,000 home sensors, with red zones indicating areas with the most distorted power.
Bloomberg News analyzed data from about 770,000 Ting sensors from Whisker Labs, which are plugged into homes across the country, to better understand the distribution and severity of an important power-quality measure known as total harmonic distortion (THD). A lower THD is better.
Good large source of data, but possibly misleading about the severity of the problem (as well as the source being somewhat dubious as it’s from a private company).
I’m from Washington and was actually surprised at how small the problem is in the Seattle area and surrounding compared to the rest of the country. We have explosive data center growth here that seems ill represented by this map.
Further, a lot of the massive data centers in Washington are actually on the Eastern side of the state, particularly in Wenatchee, which on this map is just basically entirely black. The small line of spots on the East side of the state seems more in line with Yakima/Tri-Cities/Spokane while not really including the more rural Wenatchee/Chelan area. I wish you could zoom in more on this map so I could do a proper overlay to see what areas are being missed.
Is that because it’s mostly rural and not a lot of the rural residents have the money to be adding home-sensors to be testing their energy and whether its “clean?” Like seriously, that seems more like a wealthy-people service, I had never heard of Whisker Labs or Ting before now. So not only is the data going to be limited to bigger cities (so like so many maps its really just a fucking population map), but it’s going to miss every area that isn’t as wealthy.
So Wenatchee is sparsely populated, shows up as basically black on the map, but is also home to some of the largest bitcoin mining datacenters in the state, if not the largest. Part of the reason they set up there is the cheap electricity due to close proximity to hydroelectric power. Because the population is small, more rural, and generally poorer, there’s fewer sensors showing higher THD in the area.
So anyway, a lot of words to say that this problem may be even more serious than this map shows, because there’s a lot that this map isn’t showing including the explosion of data centers in more rural areas with cheap electricity, where there not be as many rich folks with Ting sensors.
Power use by the Washington/Oregon data center cluster was almost entirely covered by a local surplus of hydropower until a couple years ago. That might be why it looks different from elsewhere.
I was going to point out that Seattle’s electricity usage is small on the map but the datacenters are going to Wenatchee and Quincy. The state has plans to remove dams and switch more to wind but the massive investment in datacenters for AI is going to derail that.
All of the data centers in the US combined use 4% of total electric load.
This added zero to the discourse other then making you look like a complete idiot. Are we in the shitpost community here?
Did you reply to the right thread mate
It was a reply to a user just ranting… Not sure what happened there 🤔
Once again, not the faulty of the technology. Don’t blame your shitty infrastructure on ai
If you’re doing a massive load increase, build out emissions-free generation to match. Some mix of wind, solar, batteries, nuclear, and geothermal would do fine. Otherwise, don’t do the big load increase.
As much as I think this is a great solution and should be written into law, the anti-ai crowd only asks it from one industry and it’s a clear sign of bias.
Not to mention that the big companies are literally doing it, either building new nuclear plants or restarting old ones. They aren’t the one holding green energy back, the oil cartel and their corrupt politicians are.
One industry? People are so mad at AI because it’s just another industry, a new one with massive environmental impact, and basically no real use outside of generating misinformation and stealing from artists. It’s the absolute worst face of the tech sector, and totally deserving of all the hate it receives.
and basically no real use outside of generating misinformation and stealing from artists
This shows you think all AI are LLMs or generative art. Those are only the most visible faces of the tech, and you’re showing your name ignorance of the field.
If you want to talk about machine learning in general, that’s a different conversation. Like it or not, colloquially, AI is LLMs and chatbots
How exactly is the rest of AI a different conversation??? Were talking about the power requirements of running AI at scale and somehow you think it’s not only correct but implied that this convo should just be about colloquial parts AI and anything else is a totally different topic in regards to power consumption?
totally deserving of the hate it gets
Yeah so breakthroughs in chemistry and other sciences for example, deserving of hate eh?
Nothing good comes from AI… when all you know about AI is colloquial lmao
Here’s an article from the IEEE about the issues with AI energy consumption. This article specifies that they’re discussing the requirements of LLMs and new generative AI. The article we’re commenting under wasn’t that cut and dry about it, but a basic understanding of the context of the world should be all a reasonable person needs to figure that out
Sorry, but you’re just wrong. Every industry is not currently spinning up their own LLM. They ARE looking to incorporate AI into their work flows, causing huge demand for data centers.
This is Bloomberg, a business centered media site. They’re not dealing with what the plebs colloquially mean.
basically no real use
"The horse is here to stay, but the automobile is only a novelty — a fad.”
Also, will you get mad at the next new industry? I highly doubt it.
If the next new industry is an energy hungry propaganda machine, yes I will
People advocating for the 99 shitty technologies that die always seem to like to quote the people talking about the one technology that survived from past generations as if that somehow made criticism of the 99 others a bad call.
Ai is going nowhere mate, you’re on the wrong side of this one. It’s too broadly useful already and has too much potential in the future
Oh fuck off with “stealing from artists” - that just proves you know nothing about the subject and and should be completely ignored.
The environmental movement asked the same thing for subsidized hydrogen production
You’re an abusive mod
More renewable energy is good, that much I will agree with
Renewables won’t exactly help harmonics, on the contrary, especially solar. This is an issue of insufficient mitigation mechanisms, probably on the supply side as computer PSUs are generally quite well-behaved loads: Drawing lots of electricity, on its own, does not harmonic distortion make.
If it is on the consumer side utilities need to start charging commercial customers for distortions just like they’re charging for blind current. If it’s on the supply side, utilities need to require large solar installations to have proper filters, and have their own mechanisms to mop up the rest. Generally the US should start having a not shoddy electricity grid, brown- and blackouts and you call yourself a developed country? We don’t even have a (colloquial) word for brownout over here!
That all said, yeah the AI hype gotta stop. That doesn’t mean that you should blame them for everything.
Generally the US should start having a not shoddy electricity grid, brown- and blackouts and you call yourself a developed country?
I just want to say I’ve never had a brownout in my part of the US, and the only blackouts we’ve had are due to weather or a car hitting a pole or something. And our electricity is inexpensive.
I’ve mostly heard of these issues in California and Texas, because of unique issues with their power utilities.
And yeah, I think both the AI hype and disdain are stupid. It’s a tool that does less than proponents claim and more than detractors claim. Don’t blame all our problems on it, and don’t suggest it’ll solve all our problems.
I agree with your idea but granting a utility the right to determine to whom they distribute power is not an easy task nor should it be taken lightly. In order to do that, you have to have regulators make the rule and then utilities obey. Utilities can’t (and shouldn’t) just deny a customer service because they don’t agree with what the customer is going to do with that power. Sanctioned natural monopolies come with regulations in most places. And in order to enforce rules, the wheels of regulatory bodies must churn and we know how slow that can be.
In theory, if you got an entity to bring x megawatts of renewable capacity online as a requirement of a new electric service load, you could tie production to data center use. Then if you ensured that the customer had controllable load to match the output of the corresponding renewable generation you could have a minimal impact growth. But that’s an absurdly complicated solution that would likely take a decade to develop and implement even if you had the political will.
I do not know what the best solution is other than to make more renewable electricity and store it, and maybe nuclear (if it didn’t take 10 years to build a plant).
I’ll go the opposite way. The fact that there are serious plans to spin up nuclear reactors to run nothing but AI datacenters is ridiculous.
Honestly I’d take the utilitarian approach to that, if it’s a net good, then I’m probably for it - but that’s a big if.
Real big if. There’s reason to believe that current models aren’t going to get much better. They’ve eaten all the training data they possibly can. Improving with further training takes exponentially more power to get a small improvement. We’re talking about new nuclear reactors because that’s what they need to get anywhere, but it’s still not going to improve by much.
The field needs a new model that can get better results on less data and less training. Then we wouldn’t need those nukes. It doesn’t appear we’ll get much better any other way.
New architechtures are in development and many have already been released. Learn something about the subject before spewing shite
You sound like the people who assured me that I needed to understand NFTs or I’d get left behind. Actually, were you one of them?
No he’s right, AI news is out of date nearly as quickly as it’s written, I’ve never seen a faster moving piece of tech.
No because nfts were obviously stupid if you had half an understanding of the technology, whereas ai is only stupid if you don’t understand the technology
Is that actually being suggested? My understanding is that only a portion of the electricity production will go to data centers in most cases, with much of the rest going to local communities. Microsoft is buying all of 3 mile island’s power, but that’s going to data centers, which do a lot more than AI.
If you’re not willing to engage in good faith, intelligent, discussion, please consider leaving the platform and making it a better place for the rest of us.
“I couldn’t be assed to read the article nor understand the problem, but I will assert my God given right to an ignorant opinion on it regardless”
Isn’t what this platform needs.
Everything gets disingenuously blamed on ai and it’s all bullshit that’s either not a problem or is a completely different problem. I don’t need to read this article to know that this problem I have read about elsewhere is not actually a problem with ai. Sorry to not reinforce your preconceived notions, you luddites are about as intractable as trump supporters and very nearly as dumb. THAT is not what this platform needs.
