Sign In or Create an Account.

By continuing, you agree to the Terms of Service and acknowledge our Privacy Policy

Technology

What Does OpenAI’s New Breakthrough Mean for Energy Consumption?

Why the new “reasoning” models might gobble up more electricity — at least in the short term

A robot with a smokestack coming out of its head.
Heatmap Illustration/Getty Images

What happens when artificial intelligence takes some time to think?

The newest set of models from OpenAI, o1-mini and o1-preview, exhibit more “reasoning” than existing large language models and associated interfaces, which spit out answers to prompts almost instantaneously.

Instead, the new model will sometimes “think” for as long as a minute or two. “Through training, they learn to refine their thinking process, try different strategies, and recognize their mistakes,” OpenAI announced in a blog post last week. The company said these models perform better than their existing ones on some tasks, especially related to math and science. “This is a significant advancement and represents a new level of AI capability,” the company said.

But is it also a significant advancement in energy usage?

In the short run at least, almost certainly, as spending more time “thinking” and generating more text will require more computing power. As Erik Johannes Husom, a researcher at SINTEF Digital, a Norwegian research organization, told me, “It looks like we’re going to get another acceleration of generative AI’s carbon footprint.”

Discussion of energy use and large language models has been dominated by the gargantuan requirements for “training,” essentially running a massive set of equations through a corpus of text from the internet. This requires hardware on the scale of tens of thousands of graphical processing units and an estimated 50 gigawatt-hours of electricity to run.

Training GPT-4 cost “more than” $100 million OpenAI chief executive Sam Altman has said; the next generation models will likely cost around $1 billion, according to Anthropic chief executive Dario Amodei, a figure that might balloon to $100 billion for further generation models, according to Oracle founder Larry Ellison.

While a huge portion of these costs are hardware, the energy consumption is considerable as well. (Meta reported that when training its Llama 3 models, power would sometimes fluctuate by “tens of megawatts,” enough to power thousands of homes). It’s no wonder that OpenAI’s chief executive Sam Altman has put hundreds of millions of dollars into a fusion company.

But the models are not simply trained, they're used out in the world, generating outputs (think of what ChatGPT spits back at you). This process tends to be comparable to other common activities like streaming Netflix or using a lightbulb. This can be done with different hardware and the process is more distributed and less energy intensive.

As large language models are being developed, most computational power — and therefore most electricity — is used on training, Charlie Snell, a PhD student at University of California at Berkeley who studies artificial intelligence, told me. “For a long time training was the dominant term in computing because people weren’t using models much.” But as these models become more popular, that balance could shift.

“There will be a tipping point depending on the user load, when the total energy consumed by the inference requests is larger than the training,” said Jovan Stojkovic, a graduate student at the University of Illinois who has written about optimizing inference in large language models.

And these new reasoning models could bring that tipping point forward because of how computationally intensive they are.

“The more output a model produces, the more computations it has performed. So, long chain-of-thoughts leads to more energy consumption,” Husom of SINTEF Digital told me.

OpenAI staffers have been downright enthusiastic about the possibilities of having more time to think, seeing it as another breakthrough in artificial intelligence that could lead to subsequent breakthroughs on a range of scientific and mathematical problems. “o1 thinks for seconds, but we aim for future versions to think for hours, days, even weeks. Inference costs will be higher, but what cost would you pay for a new cancer drug? For breakthrough batteries? For a proof of the Riemann Hypothesis? AI can be more than chatbots,” OpenAI researcher Noam Brown tweeted.

But those “hours, days, even weeks” will mean more computation and “there is no doubt that the increased performance requires a lot of computation,” Husom said, along with more carbon emissions.

But Snell told me that might not be the end of the story. It’s possible that over the long term, the overall computing demands for constructing and operating large language models will remain fixed or possibly even decline.

While “the default is that as capabilities increase, demand will increase and there will be more inference,” Snell told me, “maybe we can squeeze reasoning capability into a small model ... Maybe we spend more on inference but it’s a much smaller model.”

OpenAI hints at this possibility, describing their o1-mini as “a smaller model optimized for STEM reasoning,” in contrast to other, larger models that “are pre-trained on vast datasets” and “have broad world knowledge,” which can make them “expensive and slow for real-world applications.” OpenAI is suggesting that a model can know less but think more and deliver comparable or better results to larger models — which might mean more efficient and less energy hungry large language models.

In short, thinking might use less brain power than remembering, even if you think for a very long time.

Blue

You’re out of free articles.

Subscribe today to experience Heatmap’s expert analysis 
of climate change, clean energy, and sustainability.
To continue reading
Create a free account or sign in to unlock more free articles.
or
Please enter an email address
By continuing, you agree to the Terms of Service and acknowledge our Privacy Policy
AM Briefing

DAC Hubs May Be DOA

On Trump’s coal woes, NEPA reform, and Japan’s nuclear plans

A Climeworks facility.
Heatmap Illustration/Climeworks

Current conditions: In the Atlantic, the tropical storm that could, as it develops, take the name Jerry is making its way westward toward the U.S. • In the Pacific, Hurricane Priscilla strengthened into a Category 2 storm en route to Arizona and the Southwest • China broke an October temperature record with thermometers surging near 104 degrees Fahrenheit in the southeastern province of Fujian.

THE TOP FIVE

1. Energy Department looks ready to cancel direct air capture hubs

The Department of Energy appears poised to revoke awards to two major Direct Air Capture Hubs funded by the Infrastructure Investment and Jobs Act in Louisiana and Texas, Heatmap’s Emily Pontecorvo reported Tuesday. She got her hands on an internal agency project list that designated nearly $24 billion worth of grants as “terminated,” including Occidental Petroleum’s South Texas DAC Hub and Louisiana's Project Cypress, a joint venture between the DAC startups Heirloom and Climeworks. An Energy Department spokesperson told Emily that he was “unable to verify” the list of canceled grants and said that “no further determinations have been made at this time other than those previously announced,”referring to the canceled grants the department announced last week. Christoph Gebald, the CEO of Climeworks, acknowledged “market rumors” in an email, but said that the company is “prepared for all scenarios.” Heirloom’s head of policy, Vikrum Aiyer, said the company wasn’t aware of any decision the Energy Department had yet made.

Keep reading...Show less
Blue
Politics

How a Children’s Hospital Became Collateral Damage in the Government Shutdown

Last week’s Energy Department grant cancellations included funding for a backup energy system at Valley Children’s Hospital in Madera, California

Valley Children's Hospital.
Heatmap Illustration/Valley Children's Healthcare, Getty Images

When the Department of Energy canceled more than 321 grants in an act of apparent retribution against Democrats over the government shutdown, Russ Vought, President Trump’s budget czar, declared that the money represented “Green New Scam funding to fuel the Left's climate agenda.”

At least one of the grants zeroed out last week, however, was supposed to help keep the lights on at a children’s hospital.

Keep reading...Show less
Blue
Podcast

How China’s Power Grid Really Works

Rob and Jesse break down China’s electricity generation with UC San Diego’s Michael Davidson.

Xi Jinping.
Heatmap Illustration/Getty Images

China announced a new climate commitment under the Paris Agreement at last month’s United Nations General Assembly meeting, pledging to cut its emissions by 7% to 10% by 2035. Many observers were disappointed by the promise, which may not go far enough to forestall 2 degrees Celsius of warming. But the pledge’s conservatism reveals the delicate and shifting politics of China’s grid — and how the country’s central government and its provinces fight over keeping the lights on.

On this week’s episode of Shift Key, Rob and Jesse talk to Michael Davidson, an expert on Chinese electricity and climate policy. He is a professor at the University of California, San Diego, where he holds a joint faculty appointment at the School of Global Policy and Strategy and the Jacobs School of Engineering. He is also a senior associate at the Center for Strategic and International Studies, and he was previously the U.S.-China policy coordinator for the Natural Resources Defense Council.

Keep reading...Show less