Showing posts with label Semantic property. Show all posts
Showing posts with label Semantic property. Show all posts

Thursday, November 24, 2011

364; The computational Mind

Previous lecture I introduced you to the syntactic and semantic properties of symbols. Reason was, computers work with symbols, basic symbols and complex symbols.

In the sentence "Fido is the name of my dog" you could regard "Fido" as a basic symbol and the whole sentence as a complex symbol.

Syntactic properties are derived from the symbol itself. Properties like "It has 4 characters in a specific order", "it is black on white" and so on.

To understand what "Fido" means, a semantic property you need more than the symbol itself. At least you need me pointing at my dog saying "Look, that is Fido."

Why so much focus on this distinction between syntax and semantics? That is because some people see an analogy between brain and computer or visa versa.

Philosophers of mind who endorse the computational theory of mind, have come to the view that the mind is a computer and that thinking is symbol manipulation.

Now the question "What is a computer?" A computer is a 'syntactic engine', a device which organizes and manipulates symbols on the basis of their syntactic properties

Let me give you an example. When I use my word processor to write a story on my dog, I could use the 'Find' function and type 'Find Fido'.

What Fido is, is unknown to my computer, not even important for performing the 'find'. It is just a comparing of symbols on the syntactic level.

Yet a computer seems to respect semantic properties of complex symbols too. One semantic property is 'TRUE'. A complex symbol, a statement, has a truth-value.

Take this argument:
A. All avatars are made of pixels
B. herman is an avatar
Then the computer comes with the complex symbol:
C. herman is made of pixels

You would say…look, a computer also can handle semantical properties of complex symbolism because that conclusion, typed in by nobody but generated by the computer(program), is TRUE!

Thus a computer looks to be able to rational reasoning. It 'sees' causal relations between thoughts like A and B. However, this is not the case. We have to provide the semantics.

The computer has only a symbol manipulation rule which says: if A is marked TRUE and B is marked TRUE then you can mark C as TRUE, or a rule like if all A have property X and B is an A then B has property X.

WE have to give the computer the semantics by telling the program : A = TRUE, B = TRUE. Then it applies one of the syntax rules I just stated.

And that is what the computational theory of mind holds, that all our thinking is just like a computer does, syntactic.

Putting all this together we can say that, according to the computational theory of mind ,thoughts are complex symbols with syntactic and semantic properties.

Thinking, the manipulation of thoughts, is achieved by processors which, whilst sensitive only to the syntactic properties of the thoughts involved, nevertheless respect their semantic properties. In other words,thinking is computation.

Do you know Eliza or Elbot? Well, then read this short conversation with Elbot. You can talk for hours with him.

http://www.elbot.com/

H: You are a tin can Elbot
E: Was that a compliment or an insult?
H: Neither
E: Are you trying to make fun of my artificial intelligence?
If it was neither a compliment nor an insult, tell me
what it WAS!
H: It was an objective observation, Elbot
E: Keep making observations.
One of these days they might add up to something.
H: Yes, they add up to the observation that you are a tin
can.
E: The brilliant robot author Joseph Canrod said once:
The great enemy of perfection is thought.' Think about that!
H: Joseph Canrod did not say that.
E: I'm not sure I could explain that in language
that humans could understand.
H: I can read, Elbot
E: That's nothing new! Even I once read bar codes.
H: I repeat: Joseph Canrod did not say that.
E: I prefer not to repeat myself. In this respect
I differ greatly from most human beings.
H: You really are an empty tin can.
E: Still, I would never make the sweeping generalization
of all tins being empty, would you?
H: Right.
E: Great! Agreeing with me is the first step in your evolution
into a higher being. Please accept my congratulations.

Also have a chat with Eliza:
http://psych.fullerton.edu/mbirnbaum/psych101/Eliza.htm

After that, in a next lecture we'll discuss whether computation in a computer generates a mind or not. Did I chat with a mind, named Elbot?


The Discussion

[13:28] herman Bergson: Thank you...
[13:28] herman Bergson: and Welcome Rodney :-)
[13:29] Farv Hallison: thank you professor
[13:29] Rodney Handrick: hi Herman
[13:29] Sybyle Perdide: hard stuff again Herman.. thank you
[13:29] Farv Hallison: hello Rodney, are you a tin can?
[13:29] herman Bergson: Well Sybyle...try Eliza..she wont be hard on you at all ㋡
[13:29] Rodney Handrick: hi
[13:30] Rodney Handrick: No Farv...I'm not
[13:31] Farv Hallison: you have a pixelated outline, Rodney. I am wondering if you are a hologram.
[13:31] Rodney Handrick: No
[13:31] Farv Hallison: hello Ayi
[13:32] herman Bergson: No..Rodney is one of the die hearts of th Philosophy Class for years now
[13:32] herman Bergson: diehard is it , isnt it?
[13:32] Ayi Coeur: hello all
[[13:32] Rodney Handrick: This is true...:-)
[13:32] herman Bergson: But many questions about this lecture?
[13:33] herman Bergson: Understandable...
[13:33] herman Bergson: the message is only that computers are syntactic engines...
[13:33] Mick Nerido: Talking to Elza was like talking to a therapist lol
[13:33] herman Bergson: It means...they don't deal with content, but only with the shape of symbols
[13:34] herman Bergson: Yes Mick...she is good
[13:34] Mick Nerido: It turned my questions back on me.
[13:34] herman Bergson: I had lengthy conversations with her which made 100% sense
[13:35] Sybyle Perdide: lucky one
[13:35] herman Bergson: yes of course Mick...that is the Rogerian approach
[13:35] Sybyle Perdide: she tried to escape the conversation twice
[13:35] Farv Hallison: Did she introduce any new words into the conversation?
[13:35] Rodney Handrick: I experienced the sam e thing Mike
[13:36] herman Bergson: She can Farv
[13:36] Mick Nerido: She is not a mind but a computer program designed to resemble our minds...
[13:36] herman Bergson: Oh yes Mick...
[13:36] Ayi Coeur: who is Elza?
[13:36] herman Bergson: But then you come to Artificial Intelligence..
[13:37] herman Bergson: Eliza is a computer program Ayi...
[13:37] herman Bergson: a Rogerian psycho therapist
[13:37] Ayi Coeur: ah ok:)
[13:37] Farv Hallison: as a computer program, she could be programmed to remeber previous conversations and accumulated a big dictionary of the meaning of words.
[13:37] Mick Nerido: Cheap therapy lol
[13:38] herman Bergson: Yes Farv...that is what Elbot seems to do
[13:38] Farv Hallison: a cheap program would just spew back sex words
[13:39] herman Bergson: Eliza doesn't like bad words Farv
[13:39] Farv Hallison: so Eliza isn't cheap/.
[13:39] herman Bergson: Oh no...
[13:39] herman Bergson: It is a scientific achievement
[13:39] Lizzy Pleides: in sl we have a parrot who can talk like Eliza
[13:39] herman Bergson: The name connected to it is Weizenbaum
[13:40] Farv Hallison: I thought she was Eize Dolittle.
[13:40] herman Bergson: Joseph Weizenbaum
[13:40] Rodney Handrick: I wonder how many servers are used to run Eliza
[13:41] Rodney Handrick: And lines of code?
[13:41] herman Bergson: "Computer power and human reason"
[13:42] herman Bergson: It is not a big program Rodney...
[13:42] herman Bergson: you can find a open source Java version on the net
[13:42] Rodney Handrick: really...java...I have to look it up
[13:42] Ayi Coeur: i think it's worth trying what she has to say:)
[13:42] Sybyle Perdide: but Eliza never comes to a new level of knowledge, didn't she?
[13:42] Farv Hallison: I think I saw a LSL version.
[13:43] Rodney Handrick: what is the program code called
[13:43] herman Bergson: yes...I must have it somewhere Rodney
[13:43] Farv Hallison: I saw eliza.bas
[13:43] herman Bergson: Weizenbaum wrote it in BASIC
[13:43] herman Bergson: Where Farv???
[13:43] Ayi Coeur: auw..what a lot of work
[13:44] Ayi Coeur: wb Mick
[13:44] Rodney Handrick: I'm currently taking a Stanford U course in artificial intel
[13:44] herman Bergson: Hi Mick
[13:44] herman Bergson: cool Rodney
[13:44] Mick Nerido: hit wrong button lol
[13:44] herman Bergson: lol
[13:44] Farv Hallison: Is Eliza teaching the course?
[13:45] herman Bergson: means you might be in time for the next lectures
[13:45] Rodney Handrick: lol
[13:45] Farv Hallison: I am the wrong button.
[13:45] herman Bergson: No Farv my name is herman
[13:45] Farv Hallison: metaphorically.
[13:45] Rodney Handrick: Is this it?
[13:45] Rodney Handrick: http://smallbasic.sourceforge.net/?q=node/56
[13:46] herman Bergson: I'd love to have a source code of Eliza and translate it to LSL
[13:47] herman Bergson: THANK YOU Rodney
[13:47] herman Bergson: I might make a philosophical Eliza :-)
[13:48] Ayi Coeur: :) in basic?
[13:48] Lizzy Pleides: we prefer you Herman!
[13:48] herman Bergson: no..LSL...so that she can work in SL
[13:48] herman Bergson: Don't worry Lizzy...I am still in charge here
[13:48] Sybyle Perdide: giggles
[13:48] Ayi Coeur: and he stays that way,,,i guess..
[13:49] herman Bergson: well...I guess we are done...
[13:49] Ayi Coeur: missed the whole lecture..thanks to sl
[13:49] herman Bergson: so thank you all and today especially Rodney for the URL
[13:49] Sybyle Perdide: it was brilliant, Avi ;)
[13:49] herman Bergson: class dismissed
[13:49] Rodney Handrick: sure...not a problem
[13:49] Lizzy Pleides: yes very good!

Enhanced by Zemanta