How the brain works

SIR – In principle, artificial intelligence (AI) presents existential threats (report, January 23). Computer-run systems have the potential to be dangerous and AI will take more jobs than it creates.

But AI technology is not yet truly “intelligent”. Current programs are very domain-specific – try getting a computer-driven car to fly a plane – and while artificial general intelligence has been promised for decades, we are nowhere near achieving it.

The problem is that we don’t know how intelligent systems work. Psychologists and neuroscientists, despite recent advances, do not really understand the workings of the only conscious general intelligence system we know of: the human mind-brain.

The human sciences are in a rut; the development of AI depends on our getting them out of it.

Professor Trevor Harley

Blairgowrie, Perthshire