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 Recent data from the UK shows that the country’s outdated electricity network is holding back affordable housing projects.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.
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.
If someone wants to use a vibrator that consumes an entire city's worth of yearly energy consumption each day then I'd say that they shouldn't be allowed to do that. Making excessive energy consumption prohibitively expensive goes some way towards discouraging this at least.
If someone wants to pay that much for energy and it's priced at a level that makes it sustainable, who are we to say it's not worth it?
The main argument I've seen against higher prices for things energy and water is that it would place an undue burden on low-income people, but that's one of the many problems that could be eliminated in its entirety by a universal basic income program. Even if it's just a bare-bones program that only covers the cost of an average person's water and energy needs, such a system would give everyone an incentive to conserve when possible, and it would do it without burdening people who can't afford it.
Hmm, that makes me think we could adopt a tiered pricing system for things like water. The first 100 gallons are priced at 10 cents each, then usage beyond that goes up to 50 cents each?
You could tweak the rates & threshold to make more sense -- I don't know water rates off the top of my head, and that probably varies by orders of magnitude across the entire U.S. Also, I have no idea what water usage rates look like for different types of properties. A sports stadium, an office building, an aluminum processing plant, and a SFH with a rain garden will all have really different water usage details.
All this is kind of hinting at a broader "environmental impact" measure. That gets super complicated, though.
First gen, 4k video rendering without hardware support also consumed orders of magnitude more power than any other common thing before it, good thing we banned it, instead of you know, just letting the technology mature and develop for a couple years til we had hardware chips that could do it for almost no power.
Hell, video games are less useful for the world than AI and they consume orders of magnitude more power than ChatGPT, so let's ban them too right?
Right, so the imperative to consume less power inspired the innovation needed to make it ultimately viable in the long term. Rather than people being left to consume infinite resources without a care in the world. Let's hope that the imperative to be efficient and not use all resources all the time inspires this to also becoming viably efficient rather than regulators/officials just allowing it to spin out of control.
Yes, electricity still costs companies money so at a base level they are incentivized to minimize its usage, and then on-top of that carbon taxes should be helping to cover the environmental and incentive costs for further energy reducing innovations.
If you just want to ban electricity consuming industries I don't know why you'd start with AI, which is brand new and has genuine useful value to society, and not say, something like advertising which is just an economic distortion and massive drain on society.
Did I say at some stage that I just want a flat, nuanceless ban on industries? I answered a hypothetical posed about individual's personal consumption.
AI needs reigning in for so many different reasons, energy consumption or otherwise. Its utility to society is more than counterbalanced by the dangers that it and its unregulated expansion poses to that same society. If nothing else, government and industry bodies to catch up with it and impose appropriate standards.
If I get you right, you talk of carbon offsets. And investigation after investigation finds that the field is permeated with shady practices that end up with much less emissions actually offset.
So we absolutely should pay special attention to industries that are hogging a lot of energy. Xboxes and especially vibrators spend way less energy than data centers - though again, moving gaming on PCs and developing better dumb gaming terminals to use this computing power while playing with controllers in a living room is an absolute win for the environment.
So we absolutely should pay special attention to industries that are hogging a lot of energy. Xboxes and especially vibrators spend way less energy than data centers - though again, moving gaming on PCs and developing better dumb gaming terminals to use this computing power while playing with controllers in a living room is an absolute win for the environment.
Bruh, this is flat out a lie.
No, xboxes do not use less power when they are in your house then when they are in a data center. Servers and data center computers (including the xboxs powering xcloud), are typically more power efficient when running in optimized and monitored data centers, where they are liquid cooled with heat pumps, than when running in your dusty ass house running a fan and your houses' AC to cool them.
The power consumption of video games, if you add up every console while playing them, every server running the multiplayer and updates, and every dev machine crunching away, is a massive waste of economically unproductive energy.
The person above is right. If you want to address the climate crisis, slap a carbon tax on the cost of pollution, don't artificially pick and choose what you think is worthwhile based on your gut.
Environment doesn't stop at electricity costs, it's also about manufacturing.
A simple terminal is more efficient to produce and has way longer lifespan, removing the need to update it for many, many years.
And then you can tie it either to your existing PC (which you need anyway) or cloud (which is used by other players when you're not playing, again reducing the need for components).
That's what I meant there. Generally, from an energy standpoint, gaming can absolutely be made more energy-efficient if hardware would put it as a priority. You can make a gaming machine that needs 15W or 1500W, depending on how you set it up.
Yes and manufacturing an Xbox for every single household, boxing it and shipping it to them, and then having it sit unused for 90% of the time, has a much bigger carbon cost than manufacturing a fraction of the number of Xboxes, shipping them all in bulk to the same data center, and then having them run almost 24/7 and be shared amongst everyone.
And the same thing about optimizing gaming hardware is true for AI. The new NPUs in the surface laptops can run AI models on 30W of power that my 300W GPU from 2 years ago cannot.
I feel like we went onto two very different planes here.
Sure, data centers are more efficient than a decentralized system, but the question is, to what point the limitless hogging of power and resources makes sense?
Sure, a lot of computing power goes into, say, console gaming, but that's not what I originally talked about. I talked about data centers training AI models and requiring ever more power and hardware as compared to what we expend on gaming, first of all.
And while in gaming the requirements are more or less shaped by the improvements to the hardware, for AI training this isn't enough, so the growth is horizontal, with more and more computing power and electricity spent.
And besides, we should ideally curb the consumption of both industries anyway.
Sure, a lot of computing power goes into, say, console gaming, but that's not what I originally talked about. I talked about data centers training AI models and requiring ever more power and hardware as compared to what we expend on gaming, first of all.
But they don't. Right now the GPU powering every console, gaming PC, developer PC, graphic artist, twitch streamer, YouTube recap, etc. consumer far far more power than LLM training.
And LLM training is still largely being done on GPUs which aren't designed for it, as opposed to NPUs that can do so more efficiently at the chip level.
I understand the idea that AI training will always inherently consumer power because you can always train a model on bigger or more data, or train more parameters, but most uses of AI are not training, they're just users using an existing trained model. Google's base search infrastructure also took a lot more carbon to build initially than is accounted for when they calculate the carbon cost of an individual search.
Fair enough - I just hope the advancements in AI do not outpace our capabilities in producing a better hardware for the job, and that what's left after finds a good use in other tasks.
Because otherwise it will grow more and more into a huge ecological problem.
Nope, carbon tax is different to carbon offsets. A carbon tax is intended to put an immediate financial burden onto energy producers and/or consumers commensurate to the environmental impact of the power production and/or consumption.
From a corporations perspective, it makes no sense to worry about the potential economic impact of pollution which may not have an impact for decades. By adding a carbon tax, those potential impacts are realised immediately. Generally, the cost of these taxes will be passed to the consumer, affecting usage patterns as a potential direct benefit but making it a politically unattractive solution due to the immediate cost of living impact. This killed the idea in Australia, where we still argue to this day whether it should be reinstated. It also, theoretically, has a kind of anti-subsidy effect. By making it more expensive to "do the wrong thing" you should make it more financially viable to build a business around "doing the right thing".
All in theory. I don't know what studies are out there as to the efficacy of carbon tax as a strategy. In the Australian context, I think we should bring it back. But while I understand why the idea exists and the logic behind why it should work, I don't know how that plays out in practice.
Someone else explained how a carbon tax is different than carbon offsets, but I'll go a step further and say we should be using a cap & trade system. It would go something like this (at least in my egalitarian version):
Require "carbon credits" to be spent to legally generate carbon and other greenhouse gasses (GHGs).
Have a GHG treaty where signatories collectively decide on a GHG budget, i.e. an acceptable total level of GHG emissions.
Issue an equal amount of credits to each individual such that the total amount issued equals the total GHG budget.
Let people buy and sell carbon credits in a market system.