On-Chip Learning is Missing Neuromorphic Building Block

Elisabetta Chicca, head of the bio-inspired Circuits and Systems research group at the University of Groningen, discusses pioneering work in neural chips with memristors. She highlights the challenges of on-chip learning for spiking systems and the necessity for biologically plausible algorithms. The conversation explores the shift from 2D to 3D integration, enhancing robotic functions. Chicca also touches on the ethical implications of creating brain-like structures and the ambitious potential of neuromorphic systems to revolutionize energy efficiency in artificial intelligence.