ChatGPT’s surprising environmental footprint revealed

Climate alarmists have waged war on energy-intensive cryptocurrency mining operations, but have yet to condemn large language models (LLMs) like Microsoft-backed OpenAI’s ChatGPT, which produces 16 ounces of fresh water every time a user asks a series of questions. Uses water.

According to AP News, Microsoft revealed in its latest environmental report that its global water consumption has increased by 34% (about 1.7 billion gallons) from 2021-22, which is largely due to artificial intelligence research. There is increase.

“It’s fair to say that most of the growth is due to AI,” including “its heavy investment in generative AI and its partnership with OpenAI,” said Xiaoli Ren, a researcher at the University of California, Riverside, who is developing a new process. Are. To calculate the environmental impact of ChatGPT.

Ren’s team has calculated that ChatGPT consumes 16 ounces of water for every 5 to 50 questions or prompts a user makes. “Most people don’t know about the underlying resource usage of ChatGPT,” he said.

“If you don’t know how to use a resource, there’s no way we can help conserve the resource,” Ren said.

Given the unprecedented popularity of chatbots this year, environmental problems with LLMs have emerged as massive data centers require massive amounts of water-based liquid cooling to power chatbots.

Microsoft told the AP in a statement last week that its AI research will soon have a measure on its energy and carbon footprint. “While working on ways to make large systems more efficient in both training and application,” the tech company said.

“We will continue to monitor our emissions, accelerate progress by increasing the use of clean energy to power data centers, purchase renewable energy, and achieve our sustainability goals of being carbon negative, water positive, and zero waste by 2030. Will make other efforts to meet the goals.” to continue.

OpenAI said it had given “considerable consideration” to its computing power: “We recognize that training large models can be energy- and water-intensive” and worked to improve efficiency.

Despite significant energy and freshwater use, ChatGPT and other LLMs have not faced the same scrutiny from environmental warriors as crypto mining has over the past few years.

By zerohedge.com

More top reads from OilPrice.com:

Read this article on OilPrice.com

Source: finance.yahoo.com