Explainable AI and Biological Intelligence
Stephen Grossberg · 2026
Abstract This chapter discusses how and why biologically inspired neural network models are contributing to the development of the revolutionary computational paradigm of autonomous adaptive intelligence, whose ramifications will transform all aspects of society during this century and beyond, with applications ranging from smart appliances, self-organizing language tools, and humanoid companions, to self-driving cars, automated medical diagnostics, and self-organizing mobile robots. Such developments may be expected if only because our human brains are our best examples of systems capable of autonomous adaptive intelligence. The chapter also explains why not all neural networks can achieve these goals. Indeed, the word AI is all too often used today to signify one neural network, called Deep Learning, whose computational properties prevent it from emulating any aspect of human intelligence, for reasons that the chapter explains. The chapter also notes that AI tools like ChatGPT that build on Deep Learning literally do not know what they are talking about and would be dangerous to use in any life-ordeath application, like a financial or medical decision.