NeuroAI for AI safety

Powerful unaligned AIs pose risks to humans. This talk will explore how neuroscience-inspired AI–or NeuroAI–can lead to a deeper understanding of the human brain, and help us build more secure AI. I’ll connect these ideas to d/acc, arguing that neuroAI can play an enabling role in creating technologies that are inherently defense-favoring and promote human well-being.