@campegg Me again. I think I've muddied things up with my last reply more than made anything clearer. Let me give it another go.

Apart from the mechanical issues you mention, animals also do not posses the ability to use language propositionally. We understand that they are intelligent through different means than question and answer sessions. So when I say “detected and accepted” I’m speaking of human intelligence.

And, finally, longwindedly, I agree with the author's effort to alert the public to the dangers of assuming machines which appear to be using language actually are using language. Much peril there awaits.