A startup called Cortical Labs is pushing the boundaries of computing by building data centers that use human neurons instead of traditional silicon chips. The company's "living servers" harness brain cells to process data, achieving remarkable energy efficiency that dwarfs conventional data centers.
While artificial intelligence systems today consume gigawatts of power, the human brain operates on just 20 watts. Cortical Labs aims to replicate this efficiency by growing networks of neurons in a lab and using them as biological processors. The technology could revolutionize computing, especially for AI workloads that require massive energy.
The startup's approach involves culturing human neurons derived from stem cells and integrating them with microelectrode arrays that both stimulate the cells and read their electrical activity. This biological computing platform, which the company calls the "Biological Intelligence Operating System," learns and adapts like a living brain.
Early results show that these biological processors can perform tasks like playing computer games and recognizing patterns with far less energy than electronic chips. The next step is scaling up from small networks to server-grade systems.
"The human brain is the most efficient computer ever created," said a Cortical Labs spokesperson. "We're simply trying to harness that efficiency for modern computing tasks."
The technology has attracted significant interest from researchers and investors, who see biological computing as a potential solution to the growing energy crisis in data centers. However, questions remain about the long-term viability, ethical implications, and practicality of using living tissue for large-scale computation.