Google has released its 2024 Environmental Report , an 80-plus-page document that describes all the company’s massive efforts to apply technology to environmental issues and mitigate its own contributions. But the report avoids the question of how much energy AI consumes altogether—perhaps because the answer is “a lot more than we’d like to say.”
You can read the full report here (PDF)Frankly, this book has a lot of interesting stuff. It’s easy to forget how many dishes a large company like Google has going on, and there’s some really noteworthy work in this book.
For example, he has been working on Water Renewal ProgramThe company seeks to offset the water used in its facilities and operations, ultimately creating a net positive. It does this by identifying and funding watershed restoration, irrigation management, and other work in this area, with dozens of such projects around the world funded at least in part by Google. The company has achieved 18% of its water consumption (by whatever definition that is here) this way and is improving every year.
The company also takes great care to highlight the potential benefits of AI for climate, such as improving irrigation systems, creating more fuel-efficient routes for cars and boats, and predicting floods. We’ve highlighted some of these benefits already in our AI coverage, and they could be extremely useful in many areas. Google doesn’t have to do these things, and many big companies don’t. So, credit where credit is due.
But then we get to the section on “Responsible Management of AI Resource Consumption.” Here, Google, which has been so sure of every statistic and estimate so far, suddenly throws up its hands and shrugs its shoulders. How much energy does AI consume? Can anyone imagine that? truly Make sure?
But that must be a bad thing because the first thing the company does is downplay the entire data center energy market, saying it only accounts for 1.3% of global energy use, and that Google’s energy use is only 10% of that — so just 0.1% of the world’s total energy is powering its servers, according to the report. Silly!
It’s worth noting that the company decided in 2021 that it wanted to reach net-zero emissions by 2030, though the company acknowledges that there is a great deal of “uncertainty,” as it likes to call it, in how that will actually happen. Especially since its emissions have increased every year since 2020.
In 2023, our total greenhouse gas emissions will be [greenhouse gas] Emissions amounted to 14.3 million tons of carbon dioxide.2H, represents 13% increase year-over-year and 48% increase compared to our base year target of 2019. This result is primarily due to increased energy consumption in data centers and supply chain emissions. As we continue to integrate AI into our products, reducing emissions may be challenging due to increased energy demand due to the intensity of AI-driven computing, and emissions associated with anticipated increases in our technology infrastructure investments.
(Emphasis mine in this and the quote below.)
But AI growth has been lost amid the above doubts. Google has this excuse for not quantifying 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 integrate AI more deeply into our product portfolio, It would not make sense to differentiate between AI and other workloads. So we focus on the entire data center metrics. Because it includes the total resource consumption (and therefore environmental impact) of AI.
“Complex and sophisticated”; “Trends probably don’t capture everything”; “Distinction…wouldn’t make sense”: This is the kind of language used when someone knows something but would really rather not tell you.
Does anyone really think that Google doesn’t have a clue how much AI training and inferences add to its energy costs? Isn’t the ability to accurately analyze these numbers part of the company’s core competency in cloud computing and data center management? It has all these other statements about how efficient its dedicated AI servers are, how it does all this work to reduce the energy required to train an AI model by a factor of 100, etc.
I have no doubt that Google is doing a lot of great green work, and you can read all about it in the report. But it’s important to highlight what it seems to be refusing to highlight: the enormous and growing cost of the energy required for AI systems. The company may not be the primary driver of global warming, but despite its potential, Google doesn’t appear to have a net advantage yet.
Google has every incentive to downplay and obscure these numbers, which even in their highly efficient, discounted state can’t be good. We’ll be sure to ask Google for more details before we find out if they get any worse in the 2025 report.