Great article Doug! I’ve appreciated the subtle irony dripping off your words perhaps even more than the effort it must have cost you to play Devil’s advocate, if only for fun…
In fact you know perfectly well that the ominous AIspeak hypothesized by you is theoretically impossible, at least as long as humans stay the way they have been ever since they climbed down the trees and started using tools. Otherwise lawyers, politicians, dictators, priests, humorists and everyone else who makes his living by juggling words would starve within a few weeks.
I believe it’s a common misapprehension to regard a language - any human language - as a monolithic entity that can be thoroughly described by a complete dictionary and a few thick grammar and syntax tomes. Being made by humans for humans, a language is actually a spontaneous aggregate of countless parts, too often discouragingly overlapping, each of them covering a different area of human communication.
On one end we have a very tiny section where each word has a univocal meaning and can be translated unambiguously with a mere lookup table. Your AIspeak actually, perfectly suited for the simplest computer or for a bureaucrat.
On the opposite end there’s a large area filled with words whose meanings we “feel” without really understanding them - not just names attached to our feelings but also words for elusive concepts such as “I”, “God”, “intelligence”, “sentient”… you name them.
Somewhere in between (or on another end, it’s a matter of taste) there’s a section overcrowded with mouth-fillers, guaranteed harmless words devoided of any meaning to be used when the brain is empty.
Then there are so many sections specifically dedicated to every conceivable profession, activity, science (real or wannabe), art, hobby and whatnot. And that’s just the beginning of a very, very long list.
It looks to me a sort of wicked miracle that today a mere machine, an intrinsically deterministic and relentlessly logical contraption, may be forced to make heads or tails of the unholy mess I tried to sketch above, to the point where it can provide almost perfect human-sounding replies as if it really understood the language, even detecting humor, metaphors and figures of speech, and responding in kind.
But in this case it’s the machine that emulates the human, not the other way around.
Thus it’s not unreasonable to expect that, as the technological sophistication level rises, the machine will become gradually more human rather than the humans more computer-like.
At least that’s my hope - occasionally frustrated by the sight of some empty-eyed youngsters “being thinked” by their smartphones, apparently unaware that there’s a whole real world around them.