Drawing a human figure using human level artificial
intelligence
Note: To make this website free to the public
please click on an ad to support my sponsors or you
can make a tax-deductable donation using Paypal
(click on the donation icon on the left).
A robot with human level
artificial intelligence is drawing a female body. In
this video, I'm trying to demonstrate how a human
robot thinks when it is drawing a picture. The video
has no audio because I wanted the viewers to focus on
the thoughts of the robot and the linear steps he has
to go through in order to draw.
At the beginning, the
robot consciously decides to draw a picture of a
female body. Next, the robot's brain selects
intelligent pathways from memory. Knowledge starts to
pour into the robot's conscious and populate
primarily 4 containers: the task container, rules
container, planning container, and identity
container. There are other minor containers, but they
are not mentioned here. These intelligent pathways
form computer programs that will process the 4
containers and output intelligence. These computer
programs inside the robot's conscious can: do tasks,
do multiple tasks, solve interruption of tasks, make
decisions, act, give common sense knowledge, extract
information, provide stereotypes for objects, provide
meaning to language, predict the future, give
step-by-step instructions to do a task, compare
information, produce logic (both induction and
deduction) and so forth.
Teachers teach the
robot a long time ago to copy a picture and draw what
he sees. In order to draw a female body, the robot's
brain has to select an "image", from memory, that it
wants to draw. His conscious extract several female
pictures that he likes and he selects one. With that
image in his mind he has to copy and draw what he
sees on a piece of paper.
The linear steps to
drawing includes the following: 1. draw stick
figures. 2. use basic shapes to represent body parts.
3. flesh out the image. 4. add shading. Because of
trial and error, some artists can bypass certain
steps. While the robot is drawing, he is looking at
the image (or images) in his mind and copying every
shape and line. When he is drawing the arm, the mind
focuses on the arm of the image. When he is drawing
the head, the mind focuses on the head.
Like I said before,
the knowledge to draw a picture is learned from
school. Teachers teach this robot how to draw
pictures. There are no pre-defined instructions in
the robot's conscious that instructs the robot to
draw a picture. Furthermore, this robot can not only
draw a female body, but all objects (this is not an
expert program).
Another fact worth
noting is that drawing 1 picture is a very simple
task. Drawing a 24-page comic book is even harder.
The robot's brain learns information in terms of a
bootstrapping process. Knowledge builds on top of
each other. This knowledge about drawing pictures is
used to make a comic book. For example, in a comic
book, artists have to draw many different objects,
like cars, humans, cities, animals, scenes,
establishing shots, machines, apartments, etc. The
artist has to basically know how to draw every scene
in a movie. Drawing a 24-page comic book is not as
simple as drawing one picture. Later on, I will make
a video on drawing one panel in a comic book.