Big Tech Says Generative AI Will Save the Planet. It Doesn’t Offer Much Proof
But it turns out that many of these claims have very little, if any, actual evidence.
Joshi is the author of a new report, released Monday with the support of several environmental organizations, that attempts to quantify some of the most high-profile claims about how AI will save the planet. The report examines more than 150 claims made by technology companies, energy associations and others about how “AI will provide a net climate benefit.” Joshi’s analysis reveals that only a quarter of these claims were supported by academic research, while more than a third publicly cited no evidence.
“People make claims about what kind of societal impacts AI has and what effects it has on the energy system. These claims often lack rigor,” says Jon Koomey, an energy and technology researcher who was not involved in Joshi’s report. “It’s important not to take self-serving claims at face value. Some of these claims may be true, but you have to be very careful. I think there are a lot of people making these claims without much support.”
Another important topic explored by the report is what kind AI, exactly, what tech companies are talking about when they talk about AI saving the planet. Many types of AI are less power-intensive than the consumer-driven generative models that have dominated headlines in recent years and require enormous amounts of computation (and energy) to train and operate. Machine learning has been a staple of many scientific disciplines for decades. But it’s large-scale generative AI, particularly tools like ChatGPT, Claude and Google Gemini, that are the public focus of much of the infrastructure building by tech companies. Joshi’s analysis found that almost all of the claims he examined associated more traditional, less power-intensive forms of AI with the consumer-centric generative AI that is driving much of data center construction.
David Rolnick is an assistant professor of computer science at McGill University and president of Climate Change AI, a non-profit organization advocating for machine learning to combat climate issues. He is less concerned than Joshi about where the data from where big tech companies get their figures on AI’s climate impact comes from, given the difficulty, he says, of quantitatively proving impact in this area. But for Rolnick, distinguishing between the types of AI technologies that companies consider essential is a key part of this conversation.
“My problem with big tech companies’ claims about AI and climate change is not that they are not fully quantified, but that they rely on hypothetical AI that does not currently exist, in some cases,” he says. “I think the amount of speculation about what might happen in the future with generative AI is grotesque.”
Rolnick points out that, from techniques to increase network efficiency to models that can help discover new species, deep learning is already being used in myriad sectors across the world, helping to reduce emissions and combat climate change today. “However, that’s different from ‘At some point in the future this might be useful,'” he says. Additionally, “there’s a disconnect between the technology that big tech companies are working on and the technologies that are actually driving the benefits they claim to be adopting.” Some companies may tout examples of algorithms that, for example, help better detect floods, using them as AI examples to advertise their big language models — despite the fact that algorithms helping with flood forecasting don’t are not the same type of AI as a consumer-facing chatbot.
