Assignment for Aaron's Code - Design Computing Theory | ARCH 587, Assignments of Architecture

Material Type: Assignment; Class: DSGN COMUT THEORY; Subject: Architecture; University: University of Washington - Seattle; Term: Unknown 1989;

Typology: Assignments

Pre 2010

Uploaded on 03/10/2009

koofers-user-snv
koofers-user-snv 🇺🇸

10 documents

1 / 3

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
Chen-Je Huang
Response of two chapters excerpts from Aaron’s Code
These two chapters from "Aaron's Code" describe a painter's exploration of creating
paintings with computer. In the beginning, the painter, Harlod Cohen, just felt that
learning programming might refresh him intellectually. However, after he discovered that
programming involves rules, procedures, and corresponding actions which could be one
kind of presentation, he created a program called Aaron to draw by following certain
rules. He also added some source of entropy to process the program so the drawings are
not repeated. After years of exploration, Cohen externalized the drawing process and
improved the program. People could consider the style of these program-made art pieces
is similar to Harlod Cohen.
These two chapters do not describe the details of Cohen’s program. They merely describe
some general issues about artificial intelligent such as the difference between human-
made and machine-made.
My response to these chapters has two parts: First, machine–made paintings have same
meaning as real paintings. Second, if we want to talk about a higher level of drawing like
intention and creativity, machine could be a tool such as a brush.
I think that machine-made drawings are equal to hand-made ones. The reason is based on
the idea of creating a series of random number with computer. Most believe that those
random numbers generated by computer are random. In fact, computer only follow the
instruction people gave. These pseudo-random numbers only have the same statistic
distribution as real random number. We could analysis the result with a statistic method
or test if it is predictable to know the randomness. If people agree that pseudo number is
random when they have no idea about real randomness, why people object machine-made
drawing as real if they do not know it is drawed
How could a computer generate random numbers? There are two approaches to do this.
One is through pseudo-random number generators. The other is to sample some entropy
sources outside of computer.
Generators in the first method compute from an algorithm or take numbers from a pre-
pf3

Partial preview of the text

Download Assignment for Aaron's Code - Design Computing Theory | ARCH 587 and more Assignments Architecture in PDF only on Docsity!

Chen-Je Huang

Response of two chapters excerpts from Aaron’s Code

These two chapters from "Aaron's Code" describe a painter's exploration of creating paintings with computer. In the beginning, the painter, Harlod Cohen, just felt that learning programming might refresh him intellectually. However, after he discovered that programming involves rules, procedures, and corresponding actions which could be one kind of presentation, he created a program called Aaron to draw by following certain rules. He also added some source of entropy to process the program so the drawings are not repeated. After years of exploration, Cohen externalized the drawing process and improved the program. People could consider the style of these program-made art pieces is similar to Harlod Cohen.

These two chapters do not describe the details of Cohen’s program. They merely describe some general issues about artificial intelligent such as the difference between human- made and machine-made.

My response to these chapters has two parts: First, machine–made paintings have same meaning as real paintings. Second, if we want to talk about a higher level of drawing like intention and creativity, machine could be a tool such as a brush.

I think that machine-made drawings are equal to hand-made ones. The reason is based on the idea of creating a series of random number with computer. Most believe that those random numbers generated by computer are random. In fact, computer only follow the instruction people gave. These pseudo-random numbers only have the same statistic distribution as real random number. We could analysis the result with a statistic method or test if it is predictable to know the randomness. If people agree that pseudo number is random when they have no idea about real randomness, why people object machine-made drawing as real if they do not know it is drawed

How could a computer generate random numbers? There are two approaches to do this. One is through pseudo-random number generators. The other is to sample some entropy sources outside of computer.

Generators in the first method compute from an algorithm or take numbers from a pre-

calculated table. If we apply these generated numbers to a statistic analysis, the result is the same as real random numbers. In some way, the result of pseudo-random number generator might be predictable, but most computer applications use it to produce random number and many people have no idea these numbers are not random. The second method is sampling and processing from a source of entropy such as radioactive resource. The decay of the radioactive is not predictable so the result of the second method is not predictable.

Is machine’s drawing predictable? We know the drawings created by Aaron are human figures and plants. In some way, it is easier to predict. However, the drawings Cohen created with turtle at Kassel, the topic are random because of taking sonar feedback to the computer. I prefer the answer that machine’s drawing is not predictable.

What is a real drawing? I am not trying to do a research of this question here. However, if I just mimic a picture or what I saw, people will agree this drawing is real drawing and has some value associated, which possiblly could be considered as a art piece, I will say those machine drawings made by Cohen have same meaning as real art pieces.

I also want to use Jackson Pollock’s drawing technique to support my point. The technique Pollock developed is called spontaneous pouring technique. I believe that Pollock could control the pouring very well in his later works. However, his early works involves random strokes. Could we say that the early works are not real art? I guess the answer is not. It is the process of developing a new skill. The trials have their values.

Now, here comes another question. Even there are lots of random strokes in Pollock’s drawing, people might argue that Pollock has the intention to create something, but machine does not. Nevertheless, if we consider computer as a drawing tool for Cohen, Cohen and Pollock did the same thing. Cohen has the intention to add randomness to his works. The best supporting evidence for this is that people could tell the similarity between Cohen’s work in 1960s and those machine drawings. They share similar style defined as rules of operations. Cohen plays a dominant controlling role for the whole program.

I avoid the question to discuss if the machine has its own intelligent because I consider the machine as a tool for Cohen and the randomness of the program is what Cohen wanted to present. This also supports my first statement that machine drawings have the