Facebook<\/p>\n
Tweet<\/p>\n
<\/p>\n Email Link<\/p>\n \n Sign up for CNN\u2019s Life, But Greener newsletter. Our limited newsletter series guides you on how to minimize your personal role in the climate crisis \u2014 and reduce your eco-anxiety<\/a>.\n <\/p>\n \n Whether it\u2019s answering work emails or drafting wedding vows, generative artificial intelligence tools have become a trusty copilot in many people\u2019s lives. But a growing body of research shows that for every problem AI solves, hidden environmental costs are racking up.\n <\/p>\n \n Each word in an AI prompt is broken down into clusters of numbers called \u201ctoken IDs\u201d and sent to massive data centers<\/a> \u2014 some larger than football fields \u2014 powered by coal or natural gas plants. There, stacks of large computers generate responses through dozens of rapid calculations.\n <\/p>\n \n The whole process can take up to 10 times more energy to complete than a regular Google search, according to a frequently cited estimation<\/a> by the Electric Power Research Institute.\n <\/p>\n \n So, for each prompt you give AI, what\u2019s the damage? To find out, researchers in Germany tested 14 large language model (LLM) AI systems by asking them both free-response and multiple-choice questions. Complex questions produced up to six times more carbon dioxide emissions than questions with concise answers.\n <\/p>\n \n In addition, \u201csmarter\u201d LLMs with more reasoning abilities produced up to 50 times more carbon emissions than simpler systems to answer the same question, the study reported.\n <\/p>\n \n \u201cThis shows us the tradeoff between energy consumption and the accuracy of model performance,\u201d said Maximilian Dauner, a doctoral student at Hochschule M\u00fcnchen University of Applied Sciences and first author of the Frontiers in Communication study<\/a> published Wednesday.\n <\/p>\n \n Typically, these smarter, more energy intensive LLMs have tens of billions more parameters \u2014 the biases used for processing token IDs \u2014 than smaller, more concise models.\n <\/p>\n \n \u201cYou can think of it like a neural network in the brain. The more neuron connections, the more thinking you can do to answer a question,\u201d Dauner said.\n <\/p>\n<\/p>\n \n Complex questions require more energy in part because of the lengthy explanations many AI models are trained to provide, Dauner said. If you ask an AI chatbot to solve an algebra question for you, it may take you through the steps it took to find the answer, he said.\n <\/p>\n \n \u201cAI expends a lot of energy being polite, especially if the user is polite, saying \u2018please\u2019 and \u2018thank you,\u2019\u201d Dauner explained. \u201cBut this just makes their responses even longer, expending more energy to generate each word.\u201d\n <\/p>\n \n For this reason, Dauner suggests users be more straightforward when communicating with AI models. Specify the length of the answer you want and limit it to one or two sentences, or say you don\u2019t need an explanation at all.\n <\/p>\n \n Most important, Dauner\u2019s study highlights that not all AI models are created equally, said Sasha Luccioni, the climate lead at AI company Hugging Face, in an email. Users looking to reduce their carbon footprint can be more intentional about which model they chose for which task.\n <\/p>\n \n \u201cTask-specific models are often much smaller and more efficient, and just as good at any context-specific task,\u201d Luccioni explained.\n <\/p>\n \n If you are a software engineer who solves complex coding problems every day, an AI model suited for coding may be necessary. But for the average high school student who wants help with homework, relying on powerful AI tools is like using a nuclear-powered digital calculator.\n <\/p>\n \n Even within the same AI company, different model offerings can vary in their reasoning power, so research what capabilities best suit your needs, Dauner said.\n <\/p>\n \n When possible, Luccioni recommends going back to basic sources \u2014 online encyclopedias and phone calculators \u2014 to accomplish simple tasks.\n <\/p>\n \n Putting a number on the environmental impact of AI has proved challenging.\n <\/p>\n \n The study noted that energy consumption can vary based on the user\u2019s proximity to local energy grids and the hardware used to run AI models. \n Furthermore, many AI companies don\u2019t share information about their energy consumption \u2014 or details like server size or optimization techniques that could help researchers estimate energy consumption, said Shaolei Ren, an associate professor of electrical and computer engineering at the University of California, Riverside who studies AI\u2019s water consumption.<\/a>\n <\/p>\n \n \u201cYou can\u2019t really say AI consumes this much energy or water on average \u2014 that\u2019s just not meaningful. We need to look at each individual model and then (examine what it uses) for each task,\u201d Ren said.\n <\/p>\n \n One way AI companies could be more transparent is by disclosing the amount of carbon emissions associated with each prompt, Dauner suggested.\n <\/p>\n \n \u201cGenerally, if people were more informed about the average (environmental) cost of generating a response, people would maybe start thinking, \u2018Is it really necessary to turn myself into an action figure just because I\u2019m bored?\u2019 Or \u2018do I have to tell ChatGPT jokes because I have nothing to do?\u2019\u201d Dauner said.\n <\/p>\n
\n <\/a><\/p>\n
That\u2019s partly why the researchers chose to represent carbon emissions within a range, Dauner said.\n <\/p>\n