Google has released its 2024 Environmental Report, an 80-plus-page document outlining all of the giant company’s efforts to apply technology to environmental problems and mitigate its own contributions. But it completely sidesteps the question of how much energy AI uses — perhaps because the answer is “a lot more than we’d like to say.”
You can read the full report here (PDF)And honestly, there’s a lot of interesting stuff in it. It’s easy to forget how many plates a company as big as Google is pulling, and there’s some really noteworthy work here.
For example, you have been working on a water replenishment programGoogle hopes to offset the water used in its facilities and operations, eventually creating a net positive. It does this by identifying and funding watershed restoration, irrigation management, and other work in that area, with dozens of such projects around the world at least partially funded by Google. It has come to replace 18% of its water use (by whatever definition of that word is used here) this way, and is improving every year.
The company is also keen to highlight the potential benefits of AI in the climate, such as optimising irrigation systems, creating more fuel-efficient routes for cars and boats, and predicting floods. We’ve highlighted some of these things before in our coverage of AI, and they could actually prove very useful in many areas. Google doesn’t have to do these things, and many big companies don’t. So give credit where credit is due.
But then we come to the section “Responsible management of AI resource consumption”. Here Google, so confident of all the statistics and estimates so far, suddenly opens its hands and shrugs its shoulders. How much energy does AI consume? Can anyone In fact be sure?
It must be bad, though, because the first thing the company does is downplay the entire data center power market, saying it accounts for only 1.3% of global energy use, and the amount of energy Google uses is only at most 10% of that, so only 0.1% of all the world’s energy powers its servers, according to the report. A pittance!
Notably, in 2021 it decided it wanted to reach net-zero emissions by 2030, although the company admits there is a lot of “uncertainty,” as it likes to call it, about how that will actually happen. Especially since its emissions have increased every year since 2020.
In 2023, our total GHG emissions [greenhouse gas] Emissions were 14.3 million tCO2and, representing an increase of 13% year-over-year and an increase of 48% compared to our target base year of 2019. This result was primarily due to increases in data center energy consumption and supply chain emissions. As we further integrate AI into our products, reducing emissions may prove challenging due to increasing energy demands resulting from the increased intensity of AI computations and emissions associated with expected increases in our investment in technical infrastructure.
(Emphasis mine in this and in the quote below.)
However, AI growth is lost among the aforementioned uncertainties. Google has the following excuse to explain why the company is not specific about the contribution of AI workloads to its overall data center energy bill:
Predicting the future environmental impact of AI is complex and evolving, and our historical trends likely do not fully reflect the future trajectory of AI. As we deeply integrate AI into our product portfolio, The distinction between AI and other workloads will not be significant, so we will focus on metrics across the data center. as they include the total resource consumption (and therefore environmental impact) of AI.
“Complex and changing”; “trends are unlikely to fully capture reality”; “the distinction… will not be significant” – this is the kind of language you use when someone knows something but would really rather not tell you.
Does anyone really believe that Google doesn’t know, down to the penny, how much AI training and inference has added to its energy costs? Isn’t the ability to break down those numbers so precisely part of the company’s core competency in cloud computing and data center management? It has all these other claims about the efficiency of its custom AI server units, about how it’s doing all this work to reduce the energy required to train an AI model by 100x, and so on.
I have no doubt that Google is making many green efforts, and you can read all about them in the report. But it’s important to highlight what it apparently refuses to do: the huge and growing energy cost of AI systems. The company may not be the main driver of global warming, but despite its potential, Google doesn’t seem to be in a positive spot for now.
Google has every incentive to downplay and obfuscate these numbers, which even in its lean, highly efficient state can hardly be good. We’ll be sure to ask Google to be more specific before we find out if they get even worse in the 2025 report.