Future of Energy Forum panel: How AI could expand energy access
October 2, 2026
Four men in suits and casual wear pose on a stage with a "TULAN FUTURE" backdrop. Andrew Yawn
Artificial intelligence is driving an enormous increase in energy demand. But two Tulane University researchers say the technology could also help scientists discover materials capable of optimizing the electrical grid and increasing energy access for those who currently live without it.
That apparent paradox was at the center of a discussion at the third annual Tulane Future of Energy Forum, where Tulane physicist Jianwei Sun and chemist Daniel Straus joined Microsoft energy executive Darryl Willis to explore how AI can supercharge the race to develop new superconductors.
“AI will absolutely accelerate the pace of discovery and science development,” said Sun, professor of physics and engineering physics at Tulane University School of Science and Engineering.
Sun and Straus are part of a Tulane-led team selected by the U.S. Department of Energy’s Genesis Mission. The team’s goal is to develop an AI-driven approach to discovering new superconductors — materials that can carry electrical current without resistance and therefore without losing energy as heat.
Superconductors that can operate without expensive cooling equipment mean less energy is lost, and energy can be overall cheaper and more efficient. The U.S. Energy Information Administration estimates roughly 5% of electricity transmitted and distributed in the United States is lost.
In the same way that machine learning can sift through countless chess scenarios to identify the most promising moves, AI can help scientists narrow vast numbers of possible material candidates to those most worth testing.
“If we can make one superconductor that operates at room temperature, then we don’t need any sort of cooling whatsoever,” said Straus, assistant professor of chemistry at the School of Science and Engineering. “It could enable data centers and processors that give off significantly less heat using a lot less energy. We could cover the desert in solar panels and then run wires long distances to cities with no energy lost in transmission.
“It would be a major advance if anyone in the world could come up with any superconductor that operates at room temperature.”
Sun, the project’s lead researcher, said the team is building physics directly into the AI to help it make more accurate predictions. The approach combines established scientific knowledge with data from highly accurate computer calculations based on nature’s fundamental laws. This includes information about magnetism and electron spin — a property that makes electrons behave like tiny magnets. The model can then search large databases for promising materials, which are passed to Straus and other experimental researchers to make and test in the laboratory. Those results are then fed back into the model.
That cycle could dramatically narrow a search that traditionally requires large amounts of trial and error.
Straus said synthesizing and testing just one or two materials can take weeks. AI could instead help researchers determine which possibilities are worth pursuing before they begin the slow process of making them.
If that search eventually leads to the next generation of superconductors, the effects could reach far beyond the laboratory.
Approximately 655 million people worldwide live without electricity, according to the World Health Organization.
Outside of developing the energy systems of the future, AI is also being used to improve current energy systems, Willis said. Machine learning is being used to cut down on planning time for new oil wells. Preventative maintenance is now AI-assisted prescriptive maintenance, which uses data to offer guidance on proactively preventing equipment failures. Willis said he’s also seen AI implementation speed up the permitting process for nuclear reactors.
“There’s still a billion people on the planet who go to hospitals that don’t have reliable electricity, and two billion people on the planet still cook with indoor fires,” Willis said. “My team is consumed with helping to make sure everyone on the planet has access to safe, reliable, affordable and clean energy. And we believe that there’s an opportunity for us to use data to accelerate everything we’re doing.”