No, it's not a joke at all. Most good humor is downright serious! 😉

What the outcome of AI development will be and the ramifications for human society are indeed a question.

I am dubious that AI will be capable of computations so complex people can no longer understand them. Even the most intensive computations, such as Big Data and Bayesian analyses, rest on human-derived theorems and innovations.

Will AI become thus capable of original and innovative thought?

Is that even possible?

I am unconvinced.

A far more probable scenario in my mind is the atrophying of human intellect--a degeneration into "Idiocracy", to use the popular metaphor. It is not that AI thought will become too complex for humanity to understand, but that humanity will be too lazy to bother with understanding.

There would be no good endings in that scenario. Recalling the Star Trek episode "The Ultimate Computer", madness would seem inevitable.