However, it turns out that many of these claims have very little – if any – actual evidence.
Joshi is the author of a up-to-date report, released Monday with support from several environmental groups, that attempts to quantify some of the best-known claims about how artificial intelligence will save the planet. The report examines more than 150 claims from tech companies, energy associations and others about how “artificial intelligence will bring net climate benefits.” Joshi’s analysis found that only a quarter of these claims were supported by academic research, while more than a third did not publicly cite any evidence at all.
“People make claims about the kind of social impact of AI and what it will do to the energy system – and often those claims are not rigorous,” says Jon Koomey, an energy and technology researcher who was not involved in the Joshi report. “It’s important not to take self-interested claims at face value. Some of these claims may be true, but you need to be very careful. I think many people make such claims without much support.”
Another essential topic covered in the report is what Polite exactly the artificial intelligence that tech companies talk about when they talk about artificial intelligence saving the planet. Many types of AI are less energy-intensive than the generative, consumer-centric models that have dominated headlines in recent years, which require massive amounts of computing power and power to train and run. Machine learning has been a cornerstone of many scientific disciplines for decades. However, it is large-scale generative artificial intelligence – especially tools like ChatGPT, Claude and Google Gemini – that is at the center of most tech companies’ infrastructure expansion projects. Joshi’s analysis found that almost all of the claims he examined combined more established, less energy-intensive forms of AI with the consumer-focused generative AI that is largely driving the construction of data centers.
David Rolnick is an assistant professor of computer science at McGill University and chair of Climate Change AI, a nonprofit organization promoting machine learning to solve climate problems. He’s less concerned than Joshi about where Large Tech companies get their numbers on AI’s climate impact, given how complex he thinks it is to quantify impact in this field. But for Rolnick, distinguishing between the types of AI technology companies touted as imperative 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 in some cases they are based on hypothetical AI that does not currently exist,” 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 escalate network efficiency to models that can facilitate discover up-to-date species, deep learning is already being used in countless sectors around the world, helping to reduce emissions and fight climate change right now. “But that’s different than saying, ‘This might come in handy at some point in the future,’” he says. Moreover, “there is a disconnect between the technology that big tech companies are working on and the technologies that actually deliver the benefits they claim to advocate.” Some companies may tout examples of algorithms that facilitate better detect floods, for example, using them as examples of AI to advertise their vast language models – even though algorithms that facilitate predict floods are not the same type of AI as a consumer-facing chatbot.