Tuesday, July 31, 2007

Anthropomorphism and Artificial Intelligence

I started to read a novel entitled "Marker" by Robin Cook. In the first chapter, he describes the sperm traveling to the fertilization of an ovum using the following adjectives or phrases:

like a group of anxious marathoners, relegating the others to short and frustrating futile lives, were lost in a form of self-sacrifice, the next ordeal for these living entities, unlucky casualties, the sperm's Holy Grail, goaded by irresistible chemical attraction, hapless haploid egg, frantically. This set me to thinking about The Chinese Room, consciousness and our tendency to anthropomorphize everything in sight.

Writers of novels want to write in a colorful way in order to entice the interest of the reader. That's fine, but it also reminds me that we have a natural(?) tendency to imagine things have a self awareness and consciousness that they clearly don't have. Obviously, sperm cells don't get "frustrated" or "frantic" when faced with an "ordeal" or "self-sacrifice." One might simply chalk it up to poetic license and have done with it, except that we all have some of this tendency and it contributes to our fallibility. For example, we attribute emotions like "happiness" to a bird twittering in a tree on a beautiful spring day. We attribute a playful mood to squirrels chasing one another. Actually, we have no notion about whether or not animals feel these emotions under such circumstances. All we know is what we would feel if we acted in an analogous way. I think this is part of the mental machinery we use to predict the behavior of other humans. The philosopher Daniel Dennett writes about this ability to predict the behavior of others as a survival characteristic that has influenced the genetic wiring of the brain.

When Turing said that if a machine in a black-box responds in a fashion that would lead us to believe there's a human in the box, then the machine in the box is intelligent, he was really saying something about our limitations. We observe behavior of either black boxes or humans and are still incapable of determining whether they feel what we feel, perceive what we perceive, or know what we know in the same way we do in our own minds. It is only because we see humans as biologically the same as ourselves that makes us fairly certain that they are intelligent in the same way that we are. It is in our nature to make that assumption in order to make others actions appear predictable and rational. We have little control over that tendency. The proof is our need to anthropomorphize in the way that Robin Cook does. The more physically similar a creature is to a human, the more we credit it with "intelligence" until it's action proves otherwise. Perhaps this is why men see women as irrational rather than simply different. Perhaps this is why L-minds and C-minds see each other as irrational. It's no surprise to me that people like Searle will not be satisfied that a machine can be intelligent. Since intelligence can only be surmised from observtion of action, it's a totally optional label. It's in the eye of the beholder.


Stu Denenberg said...

From: Stu
I'm writing these comments as I read Joel's post so that I don't forget them after I've read the whole thing so it won't be very organized.

My initial response to Cook's descripton of the sperm's adventure was how we humans so desparately want to make order from chaos. We are constantly filling in the blanks and making connections --- even when there may be no connections to be made...E.M. Forster was right on when he summed up our salvation as, "Only Connect"

Joel comments: Obviously, sperm cells don't get "frustrated" or "frantic" when faced with an "ordeal" or "self-sacrifice."

And this reminds me of the story of the two Zen monks crossing the bridge. One stops and points down into the water. "Look at the fish", he says, "how happy they are!". "What!", replies the other monk, "How can you say such a thing? How do you know if the fish are happy or not?". Smiling, the first monk replies, "How do you know that I don't know?"

Now let me back up and say that I agree with Joel's contention that "Since intelligence can only be surmised from observation of action, it's a totally optional label. It's in the eye of the beholder."

But I also believe in the interconnectedness (and therefore interdependence) of everything and of the possibility that we can "know" each other --- perhaps not in a fully rational sense but in an intuitive and experiential way. I think we can agree that there are limits to rational thought.

Suzuki, another Zen master ,said that rational thinking can only take us to the realization that rational thinking is not enough to get at absolute truth. He and all mystics claim that to achieve enlightenment and absolute truth the rational, discursive mind must be abandoned in the same way we leave behind the boat that got us across the river after we reach the shore.

What do you think?


Ira Glickstein said...

From: Ira, Re: Anthropomorphism.

Thanks Stu and Joel for administering the Blog and keeping it going in my absence!

Joel's posting and Stu's response are excellent examples of what I hope this Blog can aspire to - rational discussion of important topics with opinions expressed in a collegial way rather that to "prove a point". [NOTE: While on the Loire River barge I read a copy of "The Guardian" a respected British newspaper and they put extraneous punctuation *outside* the quote marks. As this makes absolute logical sense to this C-mind, and is also the way it *must* be done when programming a computer, I have decided to adopt that style of using quote marks from now on in this Blog. If any US English teachers and editors object, let them speak their piece!]

As both Stu and Joel agree, we attribute emotions like our own to all "normal" humans we encounter on the basis of biological similarities. No one is *exactly* like anyone else, even his or her "identical" twin, but humans are sufficiently similar to each other that we confidently assume a person viewing a beautiful scene while smiling or laughing is "happy". Why not a bird in a similar circumstance? Well, birds are not human. Right, birds are not human, but they are composed of biological cells quite similar to our own, with brains and nervous systems and so on, so why couldn't they be "happy"? They are not happy in a human way, but who is to say they are not happy in a bird way?

As for Joel's objection to the use of the words "anxious marathoners" and "short and futile lives" for competing sperm, who is to say they are not?

Stu's story of the Zen monks reminded me of Garrison Keilor's favorite joke: Two penguins are standing on an ice floe and one says to the other: "You look like you are wearing a tuxedo." The second penguin replies, "Who's to say I'm not?"

The point is, if a bird appears to be "happy" who is to say it isn't?

Daniel Dennet got into a famous controversy with Thomas Nagel on "What's it like to be a bat?" (Text at: http://members.aol.com/NeoNoetics/Nagel_Bat.html) Although Nagel accepts that a bat has "experiences" he claims we humans can never really extrapolate what it is "like" to fly in the dark and "see" the world by echo-location alone. Dennet takes the opposite view. In his "intentional stance" work (see http://en.wikipedia.org/wiki/Daniel_Dennett) he claims we can imagine what it is "like" (for instance, by closing our eyes and listening to the echos of our moving feet in a hallway).

Back to the sperm and the egg! Certainly they *are* in a race. The one with the best combination of physical strength and luck will fertilize the egg, passing on the genes it carries, while the others will go into the dustbin of history, un-remembered. Since sperm have no brains or central nervous systems, I know they don't have anything like human anxiety as individuals. However, in the larger picture, they, like those of us competing in the human "rat race", are contributing to the eventual outcome of the history of the biosphere. If we, as collections of biological cells, can feel genuine emotions, who is to say that communities of sperm cells and egg cells and so on cannot feel something like anxiety or happiness -- not at the individual human level, but at a different level?

Who is to say that competing human schools of thought, or human civilizations, are, at a level higher than individual humans, not feeling something like happiness or anxiety? Not at the individual human level, but at a different level?

Ira Glickstein

PS to Joel: Searle does not totally dismiss the idea that an electronic computer can actually be intelligent. He merely says (and I agree) that a human-programmed computer of the type we have today cannot really have what he calls "strong" intelligence. We have complete knowledge of the hardware and software in a computer and there is no place for intelligence there. On the other hand, he (and I) are open to the possibility some type of copper and silicon electronic device could have "stong" intelligence.