It is peculiar to see large parts of today's AI vanguard fervently resurrecting the "cognitive architecture" frameworks of the 1970s, a forgotten chapter for many. Will things follow a similar trajectory as Japan's "fifth-generation computing" project toke in the 1980's?
