Artificial intelligence is powerful—but energy-hungry. Modern AI systems, from recommendation engines to chatbots like ChatGPT, require massive computing resources, relying on sprawling data centers that consume gigawatts of electricity. Compare that to the human brain, which operates on just 20 watts—less than a standard lightbulb.

So, what if AI could be as energy-efficient as the brain?

That’s exactly what Suin Yi, assistant professor at Texas A&M University, and his research team are working on. They’ve developed a breakthrough concept called Super-Turing AI, designed to function more like the human brain—by learning and remembering at the same time, just as we do.

Rethinking AI’s Energy Problem

Current AI models are built on architectures that separate memory (data storage) from learning (training). That means huge amounts of data have to be constantly moved across hardware, eating up time, power, and resources.

Yi points out the stark contrast: “Our brains run on 20 watts. AI data centers run on billions. That’s not scalable—or sustainable.”

As the demand for AI grows, so do the concerns: higher energy costs, more emissions, and the need for entirely new data infrastructure. Super-Turing AI offers a promising alternative.

Nature as the Blueprint

Inspired by how the brain operates, Yi’s team looked to neuroscience for answers. In our brains, learning and memory happen in the same place—at the synapse, where neurons strengthen or weaken their connections based on experience. This is called synaptic plasticity.

Instead of using traditional backpropagation—an effective but energy-intensive training method—Super-Turing AI applies Hebbian learning and spike-timing-dependent plasticity, techniques that mimic real neural behavior. Simply put: “neurons that fire together, wire together.”

These biologically inspired processes allow AI to learn on the fly, adapting as it encounters new information—without needing enormous processing power.

Real-World Results

To test the idea, researchers built a circuit based on these principles and installed it on a drone. The result? The drone successfully navigated a complex environment without prior training, learning and adapting in real time—with less power than traditional AI systems.

The Future of Sustainable AI

Super-Turing AI could revolutionize how we think about artificial intelligence. As companies push for bigger, more capable models, their ability to scale will hit hard limits unless the hardware and energy problems are solved.

“AI isn’t just about software—it’s about hardware too,” Yi says. “Without sustainable computing systems, even the smartest AI won’t be viable in the long run.”

The team at Texas A&M believes that by redesigning AI to reflect the brain’s efficiency, we can unlock the next era of intelligent systems—ones that are both smarter and greener.

“We’re building AI that doesn’t just impress—it endures,” Yi says.
“And we believe Super-Turing AI is the path to get there.”

Source