Categories

Friday, January 10, 2025

Generative AI and climate change are on a collision course

Share

In 2025, artificial intelligence and climate change, two of the biggest social disruptors we face, will collide.

The summer of 2024 broke the record for the hottest day on Earth since data collection began, sparking widespread media coverage and public debate. This also happens to be the year in which both Microsoft and Google, two of the leading immense technology companies investing heavily in artificial intelligence research and development, missed their climate goals. While this case also made headlines and sparked outrage, the impact of AI on the environment is still not widely known.

In fact, the current “bigger is better” AI paradigm – epitomized by tech companies’ drive to create ever bigger, more powerful, immense language models that are presented as the solution to every problem – comes at a very significant cost to the environment. These range from generating colossal amounts of energy to power the data centers that run tools like ChatGPT and Midjourney, to the millions of gallons of fresh water pumped through these data centers to keep them from overheating, and the tons of scarce earth metals needed to build the hardware they contain .

Data centers are already taking advantage 2% of the world’s electricity. In countries like Ireland, this figure reaches one in five of the electricity generated, prompting the Irish government to announce an effective moratorium on modern data centers until 2028. While much of the energy used to power data centers is officially “carbon neutral,” it is based relies on mechanisms such as renewable energy credits, which technically offset the emissions resulting from generating that electricity, but do not change how it is produced.

Places like Data center alley‘ in Virginia are powered primarily by non-renewable energy sources such as natural gasand energy suppliers are delaying phasing out coal-fired power plants to keep pace with changes increased demands technologies such as artificial intelligence. Data centers draw huge amounts of freshwater from restricted aquifers, pitting local communities against data center providers in various locations, including: Arizona Down Spain. IN Taiwanthe government has chosen to allocate precious water resources to chip plants to stay ahead of growing demand, rather than allowing local farmers to employ it to water crops amid the worst drought the country has seen in more than a century.

My latest research shows that transitioning from older standard AI models – trained to perform a single task, such as answering questions – to modern generative models can take up to 30 times more energy just for answering the exact same set of questions. Tech companies, which are increasingly adding generative AI models to everything from search engines to word processing software, also don’t disclose the cost of broadcasting these changes – we still don’t know how much energy is used when talking to ChatGPT or generating a photo with the Google Twins.

Much of Gigantic Tech’s discourse on the environmental impact of AI has gone two ways: either it’s not actually a problem (according to Bill Gates) or an energy breakthrough will come and magically fix everything (according to Sam Altman). What we really need is more transparency about the environmental impact of AI through voluntary initiatives such as: AI energy star a project I am leading that would aid users compare the energy efficiency of AI models to make informed decisions. I predict that in 2025 such voluntary initiatives will begin to be enforced through legislation, from national governments to intergovernmental organizations such as the United Nations. In 2025, with more research, public awareness and regulation, we will finally start to understand it AI environmental footprint and take the necessary actions to reduce them.

Latest Posts

More News