Scientists Print Fake Neurons That Slash AI Energy Use

Scientists just figured out how to print artificial brain cells that act almost exactly like the real thing, and it could solve one of AI's biggest problems.

Researchers at Northwestern University developed a new way to create synthetic neurons using flexible electronic ink printed on polymer sheets. Unlike traditional computer chips where millions of identical components work together, these printed neurons behave like actual brain cells, each with unique firing patterns.

The secret lies in what the team didn't do. Normally, scientists burn away stabilizing polymers from electronic ink after printing. But this team left some behind, creating tiny imperfections that make each artificial neuron behave differently, just like neurons in our brains.

When electricity flows through these printed neurons, they fire voltage spikes that look remarkably like real neural signals. They don't just produce simple on-off pulses either. These devices create everything from isolated spikes to sustained firing to rhythmic bursts, mirroring the complex patterns found in biological brains.

The real test came when the team connected their artificial neurons to slices of living mouse brain tissue. The biological neurons responded and fired back, proving these synthetic signals were convincing enough to activate real neural circuits.

Why This Inspires

The brain processes information with incredible efficiency, using just 20 watts of power. Meanwhile, AI companies are now building gigawatt data centers powered by dedicated nuclear plants to meet soaring energy demands.

With just two of these printable neurons and basic circuit components, the Northwestern team produced sophisticated brain-like signaling patterns. The approach represents a fundamentally different way to build computing systems, trading billions of identical components for smaller numbers of diverse, specialized processors.

Beyond reducing AI's energy footprint, this technology could transform bioelectronic medicine and brain-computer interfaces. The researchers successfully tuned their artificial neurons to match the exact timing of biological signals, opening doors for devices that communicate seamlessly with living tissue.

While it'll take years before this technology reaches commercial scale, the team has demonstrated something remarkable: we can build electronics that think more like brains and use far less power doing it.