AI boom has caused same CO2 emissions in 2025 as New York City, report claims
Study author says tech companies are reaping benefits of artificial intelligence age but society is left to pay cost
You're viewing a single thread.
The figures show the estimated greenhouse gas emissions from AI use are also now equivalent to more than 8% of global aviation emissions. His study used technology companies’ own reporting and he called for stricter requirements for them to be more transparent about their climate impact. “The environmental cost of this is pretty huge in absolute terms,” he said. “At the moment society is paying for these costs, not the tech companies. The question is: is that fair? If they are reaping the benefits of this technology, why should they not be paying some of the costs?”
So that's actually not that much? After everybody was screaming that AI is boiling the world, 8% of global aviation emissions is kind of low. And you might hate AI, but it is really more useful than Katie from Sales getting skin cancer on a beach in Thailand or that dude getting drunk on Mallorca or whatever those billionaires are doing in their private jets
Remember they have a lot of data centres planned. This is just the start. 8% of aviation emissions is huge. We should be shrinking that number ASAP, not growing it.
This and also the models are growing and, assuming it will be achieved, AGI will require even more energy.
They won’t achieve it and they’re going to sunk cost the global economy into a nosedive.
Source: trust me bro
The same source the AI techbros are using
Is it though?
I genuinely fear you all waiting for that supreme bubble burst will be either left waiting or be extremely disappointed when it happens.
Time will tell I guess.
I hate both the AI bubble (not the science behind it) and private jets and billionaires if that makes you feel better.
Also, global aviation serves an extremely useful function. Not sure that compares to fancy code autocomplete and media generation that either invalidates digital evidence in legal courts or looks like an insult to life itself.
Also, global aviation serves an extremely useful function. Not sure that compares to fancy code autocomplete and media generation that either invalidates digital evidence in legal courts or looks like an insult to life itself.
Isn't it both? There are great use cases for global aviation (like visiting your family back home) and bad use cases (like sex tourism in a third world country). There are also great things you can do with AI and bad things.
Global aviation is generally more efficient than traveling by car is it not? Not accounting for the use of private planes.
Not really:
Short/medium-haul economy flights: roughly 240–280 g CO₂e per passenger-km. long‑distance economy flights, roughly 100–130 g CO₂e per passenger‑km Average petrol/diesel car: roughly 160–200 g CO₂e per vehicle-km (tailpipe only).
So you might be able to create a scenario where driving around in your american pickup alone produces more CO2 per km than sitting in a long-distance flight. But: If you're not sitting alone in your car and you're driving something reasonable, that flips.
And flying will cover longer distances. Yes, there are a lot of people driving to spain in the summer from the netherlands, but nobody is driving from Europe to Thailand or from New York to Bali or from London to Sidney. Flying is faster and you will cover more km. And that means that even if flying would be as efficient per km than a car, it will always be worse
Yes, I will concede to that. Supposedly LLMs help a lot in certain scientific research domains replacing tedious manual work.
The thing is, the prevalence of good vs bad scenarios are inversed between GenAI and aviation I would argue. Due to lack of legal regulation we see insane amounts of funding being given for the most greedy nefarious purposes, like the elimination of the working class or artists, privacy violations for the sake of control and literally weapons out of dystopian scifi.
It's really not the same.
As far as I can understand, that number refers to current electricity usage. So the actual energy cost is significantly higher and it's currently growing at a rapid pace.
Also what's with the whataboutism? Aviation emissions are a huge problem. We don't use it as a bar for whether something is significant or not.
And it's self-reported by the datacenters. I don't want to just say that their numbers aren't reliable, but they have a vested interest in downplaying their impact.
These are datacenters that get approval to use certain power plant styles with pollution capture technology installed that are then building more generators than approved and installing none of the pollution capture tech. So extra pollution and none of the mitigation.
it is really more useful than Katie from Sales getting skin cancer on a beach in Thailand or that...
A large chunk of air transport is also freight. And business. And regular domestic travel for people going from A to B, travel that doesn't include holidays for Kate or that drunk dude in Mallorca.
And when you look at those uses, AI is still running a pretty distant second place.
8% of total global aviation emissions doesn't put it in second place. It's not even in the top 100. I don't think it ever will be... Because building huge data centers takes years and by the time there's enough data centers to make a huge dent, the previous AI data centers will have been used to make fusion power a reality.
Today's fusion reactor designs were all made thanks to AI. The kind running in big data centers.
It takes a lot of computing power to simulate fusion reactor designs!
You mean the experimental reactors? Your comment implies that there are currently big data centers running fusion reactors to power them and that is just false.
I meant that the current reactor designs were made using AI running in big data centers with supercomputers/GPUs.
It would be higher if they counted the on-site generators.
You could close down all the AI data centres, all the crypto mining farms, all the private jets, and eliminate all the billionaires…and it would be a rounding error in global greenhouse emissions.
The vast majority of emissions come from the fact that 8 billion people exist and need food, transport, heating/cooling, and industries to support their lifestyles.
The focus should be on building towards renewable ways to support whatever we want to do. Purity tests around the worthiness of types of energy use is just a distraction.
Why not both?
Yep.
There's tons of great reasons to hate AI. Tech Bros. Spam in every nook and cranny.
...But the power/water use has always been overblown. US tech is particularly sloppy about it, setting up generators and evaporators in cities, running huge training runs without proper optimization, but thats FOMO, impatience, and Tech Bro Evangelism more than a fundamental characteristic. It doesn't have to be that way, and I'd wager they can't sustain it long.
