Artificial intelligence is essential for energy transformation—that was the consensus of many speakers during an energy policy conference at Columbia.
At the daylong event on April 12, a mix of academics and policymakers, including US Department of Energy secretary Jennifer Granholm and International Energy Agency executive director Fatih Birol, gathered to review the past decade of energy and climate policy and look ahead to the future.
During a session titled “Innovating to Net Zero: The Role of Technology in the Energy Transition,” panelists discussed recent progress in low-carbon technologies and ways to unlock future technological breakthroughs. Over the forty-minute session, the word AI came up repeatedly. When moderator Vijay Vaitheeswaran, The Economist’s Global Energy & Climate Innovation editor, asked panelists to predict which technology would have the most impact on the energy transition in 2040, half predicted it would be AI.
“AI will probably be the backbone of many, many applications for decarbonization,” said Emmanuel Lagarrigue, a partner at KKR Infrastructure. He cited managing a digital and decentralized grid, making better alternative proteins, and improved carbon footprint measurement as three opportunities where AI could have the most impact.
Nobuo Tanaka, who previously headed the International Energy Agency (IEA) and now chairs the Innovation for Cool Earth Forum, selected nuclear power as his vote, but agreed that “energy transformation and digital transformation occurs together.”
Asked about AI, he shared: “This will definitely reduce the wasteful use of energy.”
While speakers were bullish about AI, they also made clear that the technology is not a silver bullet. Rather, it must be coupled with the right data and used correctly to achieve a tangible impact.
“So often in energy we don’t have the data that’s necessary to make smart decisions,” said Ann Mettler, Vice President for Europe at Breakthrough Energy. She pointed to Italy as an example—now that the country has digitized their electricity grid, it can now be steered much more carefully and smartly.
“Unless you have the data [on your energy usage] and you can aggregate that data,” Mettler said, “you will not get the intelligence that’s part of the AI.”
