Hell, Paul, I dunno how to say this well, but to me there is such a huge difference between the ability to generate intent and the sad mimicry of it that machines do that it constitutes a qualitative leap, not just a quantitative one.

I don't have much theory to throw around on it; it strikes me that just inventorying those instances where you've seen a person really generate an intention, compared to the many instances where people or devices just chew up reactions and spit them out, will make the difference pretty clear. Pre-fabbed automated signaling systems whether in machines or in people don't carry the intent to communicate. They're displaced in time and context.

Here's a silly example -- you know those tags they put on pillows and mattresses that say "Do Not Remove under Penalty of Law"... how many times have you torn one off? :>D They have no intention, they're in the wrongplace and time as far as a communication goes -- they're meant for someone else and they are so far removed from any original intention that the communication impulse is lost in time and space.

Or when Windows starts up and says, "Welcome!". Or when you ask someone how are you, and they say, "Fine" before the words are out of your mouth, automatically.

Or when a web page says , "Congratulations, you've been selected".

These are all almost completely intention-free signals.

There are, as Bill D put it, about 47 pages of exceptions, clarifications, and so on, omitted here.

A