In contrast to current times, in the early 1970s when Goertzel first learned about AGI, it was not obvious that superhuman intelligence was feasible due to how primitive computers were; making the inference from primitive 1970s-era computers to superhuman AI required creative intellectual work and was not obviously justified.