DD - got up this morning with a cup of coffee and reread your comment. Makes perfect sense now - must have had a mild aneurism yesterday.

Some thoughts:
Like I said, I don't know the Jaynes stuff. But I'm suspecting that part of his evidence is the texts of homeric and pre-homeric times (the sumerian stuff, etc.). And I'm guessing that because there is a long tradition in Greek philology, history of religions, going back 50 so years that also attributes to this age a kind of infancy of consciousness and individual agency (cf. Bruno Snell, Dodds). And it is still orthodoxy in those fields. It is based on the way Homer and the others explain action - the causal implication of gods, the absence of deliberative internal discourse - that suggests they lacked full free will and agency, a strong sense of individual responsibility. Hence my suggestion about Bernard Williams. It's a beautiful take-down of this reading of the texts based on a lot of philosophical confusions, especially post-Kantian confusions of what it is to have a will. Williams actually goes further: he claims that Homer and his contemporaries had a much better self-understanding than us moderns who are stuck within a dogmatic and artificial conceptual frame. And by 'moderns' he means everyone from Plato onward (hahah).

As for the Turing test, this is my take. It's a sometimes useful sufficient criterion for intelligence/consciousness as applied to machine intelligence. We have a linguistic interface with them - though the language is often formal - and so, if you want to know whether they are just highly complex mechanical stimulus-response organisms or rather fully intelligent and capable of reutilizing 'mental' representations in diverse and novel circumstances - in short whether they can manipulate representations in the way we call 'thinking' and 'deliberating', then you can do it for the most part through this interface. But language is only one way in which intelligence can be observed. And it severely restricts the kinds of intelligence that can be observed. Most IT is conceived for particular compartmentalized purposes, but they already show a great deal of intelligence within those circumscribed fields. Would these computers pass the Turing test? No. Do they have a form of self-awareness? Yes, they learn, correct their own reasoning in the light of new evidence. There's a lot of use of funky new sorts of logic functionally very close to our non-deductive reasoning and learning.

I got invited to an interdisciplinary conference on AI and ethics last year (neither of this is really my field, but go figure...). I laughed at the program as I read some symposium titles having to do with Computer Rights. It seemed a mildly insane topic to worry about. It's still in its infancy, but I came out taking it all much more seriously. Right now we've only got something like the Turing test to model/assess agency; we need something sharper, and we need a better understanding of the Turing test (or at least the people in this field do).

That's all a little aside. I used the Turing test in the Mankiw piece because it's particularly relevant to the internet and these anonymous exchanges. The internet is one continuous mutual Turing test, as all we have is this little linguistic interface, and we are forced to use this restricted medium to convey thoughts and to assess thoughts of others. No facial expressions, gestures, context, background knowledge, social strata. Just brute words. And it makes it that much more fascinating. It makes us smarter, it forces us, as we have to prove every day that we aren't just pseudo-reflective automatons. As was hinted at in one of my link-love pieces, even smart people are stupid unless you force them to reason deliberately about something. We lazy mammals, when we sit down at these terminals, we are forced to crank up the frontal lobes and make an effort.

A little example: I was up late recently and saw an old film with Don Johnson playing an intellectual who had to teach a bimbo played by Melanie Griffith to get by in Washington cocktail parties. He teaches her to memorize eight lines with a lot of pundit gibberish which cover any possible topic/question she might face. She goes to the party and everyone is very impressed. Stupid movie, but the 'joke' seemed so spot on.

As for animal intelligence, it is one of those fields where we keep throwing out the facts because they don't fit the frame. And I find my frame pretty resistant as well, to be honest. Some of the behavior is pretty amazing, the complexity of animal communication, the intelligence behind the hunting techniques of dolphins (don't know the wolf stuff...), the ability of chimps to learn sign language with over a hundred terms, the tool-making skills of new caledonian crows (or robins, don't remember), the ability of elephants to do representative painting. I see and I read these things, and I am amazed and i don't know what to think. One example I was shown a clip of: a bird sitting in a lab next to a bottle with some seeds at the bottom in a little basket which he can't get at. The bird looks around and sees a stick which someone has conveniently left there. He takes the stick in his beak, slides the far end down the bottle to get at the seeds. Only they won't come out, because you need a hook to heave the basket out of there. So the bird takes the stick out again, BENDS THE END OF THE STICK to form a hook, and puts it back in and heaves those damn seeds out of there! How do I model this within some dumb stimulus-response frame? I'm suspecting there's a lot of training, by some scientist who wants to impress all us skeptics out there, that goes into this kind of behavior; TRAINING and not learning. But, yes, my frame is trembling under the strain. The frame is less worried about chimp 'language' because what we usually say here is that the chimp lacks syntax, which is what makes our language superior, or language 'proper', the kind of representational manipulation that goes into reflective thought.

Now I've forgotten where I was going with this, but maybe that is enough for now. Hope you get to see this, or I'll start pasting it in comments elsewhere...

gotta run.

my toil doth please the lady of the manor. My heart skips a beat... :0)

i better convert to presbyterianism quick...

Careful Q, he'll just starve you, pop a housing bubble on your ass and drive you into depression...

Man, I never dreamed I'd ever get this prize...! I'd like to thank ma producer, ma mama, (I love you), ma ... Don't start the damn music on me!

CCR - i tried to fit in a lame joke once on bondholders almost cutting their hair; and a damn hippie got mad at me. (crap, he's probably going to see this)

These folks seem to be in self-destruct mode.

GZ - thanks for stopping by! You're worse than DD. Now I'm going to have to go away and brush up on my chicken science...

So I'll leave that point aside for now. I'll just make two points on the robot-bigotry issue. The best of these humanoid robots are getting built in Japan by a guy called Mori. He's a slight nutter as he's built twins for each member of his real family. The facial expressions of the robots reproduce those of the originals down to margins of error within micrometers. His poor daughter, I remember hearing somewhere, was traumatized the first time she saw her twin, and I don't think they're getting along much better now. That's for the anecdote.

The more general point is that Japan has a much less adversarial relationship with robots than we do. We know our robots, for whatever reason, as primarily nasty killing machines plotting to take over the world. The Japanese like them as pets or funny dancers, and more generally represent them as human-friendly heroes of society. I don't know why this is so (all theories are welcome). So, although I haven't actually seen any of the raw data on this 'valley' phenomenon, I'm assuming a lot of these experiments on our general fowlness with the humanoid bots comes from over there. So it's not a general bot-bigotry that turns into something nastier as the threat seems so much closer.

The third of my two points is that there is a complication to the valley phenomenon. We don't want to peck them to death. It's not hatred, but unpleasant revulsion. And this revulsion has a flipside. So there are some controversial experiments going on, not on our attitudes to robots, but to virtual avatars. This is my rough memory of what they involve: It's a version of the old Millgram experiments which you learn about in most social science 101 classes [reminder: subjects are given orders by an authoritative experimenter to push a button which he is led to believe gives electric shocks to a guy behind a curtain. And, most subjects, to our general horror, ended up following the experimenters orders until the point where the subject believes the guy next door is electrocuted to death]. So now people are redoing these experiments where it's the avatar which screams and winces "in pain" every time the subject pushes the button. Now the interesting bit is that the more human-like the avatars get, the more subjects resist the order to push the button. It's bizarre, they know it's just a virtual representation! So there is a kind of reverse valley effect - an uncanny mountain, if you will. So this problem is just getting weirder.

(It's nice you guys like this stuff!)