C- The Information Machines and the Dilemma of Communication

1) Warren Weaver, “The Mathematics of Communication”, Scientific American, 1949.

Yuan:

1) Explain more about the text in the end of paragraph on page 33, “One then refines or extends…….”. What’s the author’s logic here?

2) What is the definition of repeating word “discrete” compared to “continuous”? On which method that letters are considered as discrete while speech is continuous?

3) What’s the difference of Weaver’s diagram of communication system from the one in the movie?

William

If at some point the communication between man and machine can occur at C/H (with a total, or infinitely close to total negation of engineering noise) will we have by default managed also to eliminate semantic noise?  If so, what does this imply about the nature of intelligence? If not, when/will we ever achieve it?

Rob-

1.      Why does Weaver try to quantify seemingly un-quantifiable values into an equation (pages 34-35)?

2.      When talking to another person, Weaver says that they may not get 100% of the information you are trying to convey, but you can never actually tell how much the other person has heard since you translate their response.  This seems to create a never ending loop.  How do we actually convey the information we want to?

3.      In Eddington’s quote at the end of the piece, how does he define which grouping entropy goes with (amongst distance, mass, electric force, beauty, and melody)?  Does the meaning play into the grouping?

Yi:

Q1: (P30,31) How to understand  “ information is associated with the amount of freedom of choice we have in constructing messages” ? and how it be compared with the second law of thermodynamics that entropy always increases?

Q2: (P33) why does Weaver say “this undesirable uncertainty – this noise or equivocation- will always be equal to or greater than H minus C”?

Q3: (P37) What does the author mean by saying “information and meaning may prove to be something like a pair of canonically conjugate variables in quantum theory, that is, that information and meaning may be subject to some joint restriction that compels the sacrifice of one if you insist on having much of the other.”?

2) Alan Turing, “Computing, Machinery and Intelligence”, Mind : A Quarterly review of Psychology and Philiosophy, 59(239), October 1950.

Yuan:

1) What’s Turing’s strategy on his proposal of child machine that responding the last of three components to imitate an adult human mind? (page 62)

2) According to Turing’s method that experimenters firstly create “the initial state of the mind, say at birth” and then use their judgement to process nature selection, is there a conflict that the attempt of experimenters’ cause tracing would be impossible because Turing also mentioned that “An important feature of a learning machine is that its teacher will often be largely ignorant of quite what is going on inside,” (page 63)

3) If the child machine which really rely on experimenter’s training as “Nature selection” really works out, what will be the AI’s and human’s destiny on the basis of the difference of human’s variety and the consistent “nature”. In other words, when man plays god’s role.

William

Might at some point Turing’s sweeping implications about the nature of intelligence require a very direct challenge? As machines approach human computational capacity, will we be forced to reopen the investigation of the nature of machine intelligence, and ask not “Are there imaginable digital computers which would do well in the imitation game?” but very seriously “Can machines think?” (Or perhaps by then more importantly, how do machines think?)

Rob-

1.      What is the symbol that is used throughout the piece?

2.      In section 6, Turing discusses the question of whether or not anything can really be “new” or if things are always just a product of something else.  Just because something is a product of something else doesn’t necessarily mean it’s not new.  Couldn’t it be argued then that humans, and everything else for that matter, never do or think of anything truly new?

3.      In section 8, Turing seems to be redefining “machine” so that humans can fit into the definition.  By this argument, couldn’t it be reasoned that, since machines do many things humans do (body mechanics, thought processes, etc.), the word “human” could be redefined to include machines?

Yi:

                    Q1: Is it appropriate to replace the question “Can machines think?” by the “imitation game” as Turing argued?

                    Q2: (P56) Does the author’s “argument from consciousness” persuasive with the “viva voce” example he gave?

                    Q3: (P61) In the “skin of an onion” analogy, how does Turing define the “real” mind and “the whole mind is mechanical”?

3) Charles and Ray Eames, A Communication Primer, (Film) 1953.

Yuan:

1) “Redundancy” in the film is portrayed as simply repeating “I love you, I love you, I love you”, and English language just the same as Weaver’s text. What is this term’s content while it is introduced to combat noise? Repetition or fixed structure?

2) Does annoying sound indicating if the punching machine is working functionally or not really builds a feedback or not?

3) How to define the noise and signal when talking about deciphering the nature phenomenon, such as wave, birds school and stars?

William

The Eames film describes the human brain as a hyper complex version of a digital computer, but current(?) science disputes this.  If Charles and Ray are correct, how does Turing’s dismissal of the nature of intelligence hold up? If current science is correct, does this change the nature of the Man/Machine interface?

Rob-

1.      The video is about noise and interference in communications.  However, with age, it has become grainy and wobbly, which interferes with its communication.  Isn’t this ironic?

2.      The video discusses how information is transmitted through a series of decisions (the black and white printer).  As the number of decisions increases, the information becomes clearer.  Would there be a point where there have been too many decisions so that the information becomes too dense and difficult to understand?

3.      In the discussion on waves, it is said that in order to interpret the waves, one must know how to interpret the code.  The same is true with any information transfer.  The noise and interference during the transfer has the ability to jumble the code, but the code might also be altered if the person on the other end deciphers the code incorrectly.  Is that what the video was getting at?

Yi:

              1. Should the term “noise” be defined from the sender’s perspective or the receiver’s understanding?