Bridging from natural to artificial intelligence

(Lecture Hall F3)

Despite the immense success of modern artificial intelligence, conventional computing systems cannot match biological neural processing systems in many closed-loop real-time sensory processing tasks and consume orders of magnitude more power than biological brains.
To bridge this profound efficiency and performance gap, researchers must look beyond traditional computing paradigms and draw inspiration directly from the brain's computational primitives. Exploring these cutting-edge intersections, this workshop will gather experts to discuss how biological principles can shape the future of NeuroAI.
The symposium features a keynote talk on neural architectures and local learning mechanisms, detailing the essential components needed to realize bio-inspired computing systems. This foundational overview will be followed by rapid research spotlights from outstanding postdocs covering advanced topics such as predictive coding, cortico-hippocampal planning, and bio-inspired neuromorphic computing, ultimately culminating in a dynamic open panel debate on the future of biologically grounded AI.