• Living neurons learn to carry out computing tasks automatically
• Experiment opens possibilities for merging brains with AI technology

A team of scientists successfully trained neurons taken from rat brains to perform computing tasks similar to artificial intelligence, in an experiment seen as a major step toward integrating biological brains with modern technology, according to Gizmochina.

Researchers explained that the new system connects living neurons with precise devices that record their activity, then convert it back into electrical signals that continuously feed the network, allowing the neurons to adjust their behavior on their own without direct human intervention.

To prevent over-synchronization between neurons, the team arranged them in a small lattice of tiny pores, enabling more coordinated and efficient activity.

The system was able to generate a variety of wave patterns and even mimic some of the complex behavior of chaotic systems, achieving high accuracy in computations.

Despite these successes, the system struggles when training stops, especially due to the delay in feedback signals, which makes handling fast-changing signals challenging.

Researchers hope to develop faster devices to improve performance, paving the way for future applications such as smart prosthetics, brain-machine interfaces, and hybrid bio-AI systems.

This experiment marks an exciting step toward exploring the computing potential of living brains and suggests a future where biological life and artificial intelligence could work together in unprecedented ways.