Big tech is playing its part in reaching net zero targets, but its vast new datacentres are run at huge cost to the environment, says economics professor Mariana Mazzucato
Despite its name, the infrastructure used by the “cloud” accounts for more global greenhouse emissions than commercial flights. In 2018, for instance, the 5bn YouTube hits for the viral song Despacito used the same amount of energy it would take to heat 40,000 US homes annually.
Large language models such as ChatGPT are some of the most energy-guzzling technologies of all. Research suggests, for instance, that about 700,000 litres of water could have been used to cool the machines that trained ChatGPT-3 at Microsoft’s data facilities.
Additionally, as these companies aim to reduce their reliance on fossil fuels, they may opt to base their datacentres in regions with cheaper electricity, such as the southern US, potentially exacerbating water consumption issues in drier parts of the world.
Furthermore, while minerals such as lithium and cobalt are most commonly associated with batteries in the motor sector, they are also crucial for the batteries used in datacentres. The extraction process often involves significant water usage and can lead to pollution, undermining water security. The extraction of these minerals are also often linked to human rights violations and poor labour standards. Trying to achieve one climate goal of limiting our dependence on fossil fuels can compromise another goal, of ensuring everyone has a safe and accessible water supply.
Moreover, when significant energy resources are allocated to tech-related endeavours, it can lead to energy shortages for essential needs such as residential power supply. Recent data from the UK shows that the country’s outdated electricity network is holding back affordable housing projects.
In other words, policy needs to be designed not to pick sectors or technologies as “winners”, but to pick the willing by providing support that is conditional on companies moving in the right direction. Making disclosure of environmental practices and impacts a condition for government support could ensure greater transparency and accountability.
There are layers of wrong and stupid to this article.
Despite its name, the infrastructure used by the “cloud” accounts for more global greenhouse emissions than commercial flights.
"The cloud" accounts for something like 80% of the internet across the entire planet. I'd be curious what 80% of transportation infrastructure would end being in comparison... no takers? We're only comparing to (some) flights instead of, I dunno, the vast bulk of our fossil fuel powered transport infra?
In 2018, for instance, the 5bn YouTube hits for the viral song Despacito used the same amount of energy it would take to heat 40,000 US homes annually.
Oh no, the most popular song in the world used the same amount of energy as 40k homes in the US. The US probably has something in the range of a hundred million homes. The efficiency of computing equipment increases by a sizable percentage every single year, with the odds being good the same data could be served at 1/20th the cost today. So why aren't we talking about, say, heat pumps for those homes? You know, since they're still using the same amount of energy they did in 2018?
...about 700,000 litres of water could have been used to cool the machines that trained ChatGPT-3...
Additionally, as these companies aim to reduce their reliance on fossil fuels, they may opt to base their datacentres in regions with cheaper electricity, such as the southern US, potentially exacerbating water consumption issues...
What is this idiocy? You realize that a chip fab uses something to the tune of ten million gallons of water per day, right? Ten million. Per day. I'm not even looking at other industrial processes, which are almost undoubtedly worse (and recycle their water less than fabs) - but if you're going to whine about the environmental impact of tech, maybe have a look at the manufacturing side of it.
Furthermore, while minerals such as lithium and cobalt are most commonly associated with batteries in the motor sector, they are also crucial for the batteries used in datacentres. The extraction process often involves significant water usage and can lead to pollution, undermining water security. The extraction of these minerals are also often linked to human rights violations and poor labour standards.
Man, we're really grasping at straws here. More complaining about water usage, pollution, water security, labor standards, human rights violations... wait, were we talking about the costs of data centers or capitalism in general? Because I'm pretty sure these issues are endemic, across every industry, every country, maybe even our entire economic system. Something like a data center, which uses expensive equipment, likely has a lower impact of every single one of these measures than... I dunno... clothes? food? energy production? transport? Honestly guys, I'm struggling to think of an industry that has lower impact, help me out (genuine farm to table restaurants, maybe).
There are things to complain about in computing. Crypto is (at least for the time being) a ponzi scheme built on wasting energy, social media has negative developmental/social effects, etc. But the environmental impact of stuff like data centers... its just not a useful discussion, and it feels like a distraction from the real issues on this front.
In fact I'd go further and say its actively damaging to publish attack pieces like these. The last few years I didn't drive to the DMV to turn in my paperwork, I did it over the internet. I don't drive to work because I'm fully remote since the pandemic, cutting my gas/car usage by easily 90%. I don't drive to blockbuster to pick out videos the way I remember growing up. The sheer amount of physical stuff we used to do to transmit information has been and is gradually all being transitioned to the internet - and this is a good thing. The future doesn't have to be all bad, folks.
What is this even? Batteries for UPS in a datacenter wouldn't be a patch on even a few days of production of EVs, water isn't being shipped from "drier parts of the world" to cool datacenters, and even if it were, it's not gone forever once it's used to cool server rooms.
Absolutely, AI and crypto are a blight on the energy usage of the world and that needs to be addressed, but things like above just detract from the real problem.
This is horrible article. The only number given related to LLM is 700,000 liters of water used, which is honestly minuscule in impact on environment. And then there are speculations of “what if water used in aria where there is no water”. It is on the level of “if cats had wings, why don’t they fly”.
Everything we do in modern would consumes energy. Air conditioners, public transport, watching TV, getting food, making elections… exactly the same article (without numbers and with lots of hand waving) could have written. “What if we start having elections in Sahara? Think about all the scorpions we disturb!”
So... Absolutely need to be aware of the impact of what we do in the tech sphere, but there's a few things in the article that give me pause:
Research suggests, for instance, that about 700,000 litres of water could have been used to cool the machines that trained ChatGPT-3 at Microsoft’s data facilities.
"Could". More likely it was closed loop.
Water isn't single use, so even if true how does this big number matter.
What matter is the electrical energy converted to heat. How much was it and where did that heat go?
Moreover, when significant energy resources are allocated to tech-related endeavours, it can lead to energy shortages for essential needs such as residential power supply. Recent data from the UK shows that the country’s outdated electricity network is holding back affordable housing projects.
Can you say non sequitur ?
The outdated network holding back housing is that it doesn't go to the right places with the capacity needed for the houses. Not that OpenAIUK is consuming so much that there's no power left. To use a simily, there's plenty of water but the pipes aren't in place.
This article is well intentioned FUD, but FUD none the less.
Large language models such as ChatGPT are some of the most energy-guzzling technologies of all. Research suggests, for instance, that about 700,000 litres of water could have been used to cool the machines that trained ChatGPT-3 at Microsoft’s data facilities.
Dunno about Microsoft and AWS but AFAIK Google has been powering all their data centers with "renewables" for a very long time.
I'm pretty sure many of these data centers have dedicated power sources due to the high consumption, and opt for things like hydroelectric due to cost per watt.
And at least there's a serious end product delivered, unlike crypto mining which wastes trillions of hashes to make a secure transactional network.
Despite its name, the infrastructure used by the “cloud” accounts for more global greenhouse emissions than commercial flights. In 2018, for instance, the 5bn YouTube hits for the viral song Despacito used the same amount of energy it would take to heat 40,000 US homes annually.
Mixing and matching abstract measurements doesn't work when comparing two things.
AI -and cryptocurrencies- use massive amounts of energy and the only value they produce is wealth. We don't get correct, reliable and efficient results with AI, and we don't get a really useful currency but a speculatory asset with cryptocurrencies. We are speeding into a climate disaster out of pure greed.
This is what pisses me off so much about the climate crisis. People tell me not to use my car, but then microsoft just randomly blow out 30% more co2 for AI
To no surprise, the other comments are full of laypeople that feel they understand the entire field they have never studied well enough to preach to others about just how useless and terrible it is, who also know nothing about the subject.
the infrastructure used by the “cloud” accounts for more global greenhouse emissions than commercial flights
This comparison is bad. Commercial flights don't use electricity, they use jet fuel, pumping fumes directly into the atmosphere. I don't see a single complaint about HOW electricity is produced. I just read about how there's too much solar power in California. A serious disconnect in the logic blaming AI for pollution when we should be blaming the way we produce electricity.
Love how we went from "AI needs to be controlled so it doesn't turn everything into paperclips" to "QUICK, WE NEED TO TURN THE PLANET INTO PAPERCLIPS TO GET THIS AI TO WORK!!"
So when exactly is all of this going to stop? First we had town-scale crypto farms, that were juicing enough energy to leave other people with no electricity. Then we switched to NFTs, and the inefficient ever-growing blockchain, and now we're back to square one with PISS, and it telling people to put glue on pizza, and suicide off the golden gate bridge
Of course it would.. lmao are you kidding me? Have you never seen a server farm? Hell NSA has huge warehouses of servers.
Last year, before I joined this organization, IT decided to get off Microsoft's cloud service because after some calculations they realize that on-prem hosting was significantly cheaper than cloud hosting. Now I believe more and more organizations small and large/enterprise are getting off cloud or doing a mixture / hybrid because the costs are not justifiable.
And for AI? Requiring GPUs? Huge energy consumers.
This isn't a good situation, but I also don't like the idea that people should be banned from using energy how they want to. One could also make the case that video games or vibrators are not "valuable" uses of energy, but if the user paid for it, they should be allowed to use it.
Instead of moralizing we should enact a tax on carbon (like we have in Canada) equal to the amount of money it would take to remove that carbon. AI and crypto (& xboxes, vibrators, etc) would still exist, but only at levels where they are profitable in this environment.
AI Training is a flexible energy consumer, meaning it can be switched on and off at will, so that it can take advantage of excess solar power during the daylight, providing extra income to solar panel parks. The important thing to do is to install solar panels, and then AI training isn't an environmental problem anymore.
It is hardly news that the tech bubble’s self-glorification has obscured the uglier sides of this industry, from its proclivity for tax avoidance to its invasion of privacy and exploitation of our attention span.
The industry’s environmental impact is a key issue, yet the companies that produce such models have stayed remarkably quiet about the amount of energy they consume – probably because they don’t want to spark our concern.
Google’s global datacentre and Meta’s ambitious plans for a new AI Research SuperCluster (RSC) further underscore the industry’s energy-intensive nature, raising concerns that these facilities could significantly increase energy consumption.
Additionally, as these companies aim to reduce their reliance on fossil fuels, they may opt to base their datacentres in regions with cheaper electricity, such as the southern US, potentially exacerbating water consumption issues in drier parts of the world.
In an era where we expect businesses to do more than just make profits for their shareholders, governments need to evaluate the organisations they fund and partner with, based on whether their actions will result in concrete successes for people and the planet.
As climate scientists anticipate that global heating will exceed the 1.5C target, it’s time we approach today’s grand challenges systemically, so that the solution to one problem does not exacerbate another.
The original article contains 766 words, the summary contains 214 words. Saved 72%. I'm a bot and I'm open source!
New technologies will sometimes need more energy. Thats hardly news. If we continue yo switch to renewables the impact will also be small. AI isnt even listed as its own point, heck it is not even listed in most energy budgets, yet it sounds like there will be no energy left for the rest, which is laughable, since it likely uses around 1% of the energy needed (its estimated at 2% for it in general)
The ugly truth behind journalist: broke English majors are guzzling resources at planet-eating rates
By age of 21 most journalist have produced 336 metric tons of Co2 and and 20 000 lbs of waste just to produce stacks of advertorials, mild propaganda and personal vendettas leaving a trail of blog posts across multiple servers.