Bridging Brains and Machines: A Unified Frontier in Neuroscience, Artificial Intelligence, and Neuromorphic Systems

A landmark position paper authored by a consortium of 46 researchers, including Sohan Shankar and Yi Pan, argues that neuroscience, artificial general intelligence (AGI), and neuromorphic computing are rapidly converging into a single, unified research frontier. Titled 'Bridging Brains and Machines,' the survey proposes that core principles of brain physiology—such as synaptic plasticity, sparse spike-based communication, and multimodal association—provide a direct blueprint for designing the next generation of AGI systems. It traces how modern AI innovations, from transformer attention in models like GPT-4 to multi-agent architectures, already mirror neurobiological processes like cortical mechanisms and working memory.

The paper then shifts to the hardware required to realize this vision, highlighting emerging physical substrates like memristive crossbars and in-memory compute arrays. These technologies aim to shatter the von Neumann bottleneck—the traditional separation of memory and processing—to achieve the energy efficiency of the human brain in silicon. Finally, the authors outline four critical interdisciplinary challenges: integrating spiking neural networks with foundation models, enabling lifelong learning without catastrophic forgetting, unifying language with sensorimotor learning in embodied agents, and establishing ethical safeguards for advanced neuromorphic autonomous systems. This work serves as a comprehensive agenda to guide future research across these historically separate fields.