Saibal Mitra wrote:
>
>Suppose you are a virtual person, programmed by me and living in a virtual
>environment. You do some experiments to find the laws of physics. You try to
>break up things and look what they are ``made of´´. Would you ever discover
>how the pentium processor works if you proceed this way?
>
Interesting problem.
I would set up experiments to test the limit of my world. I would find
out that I get a exception interrupt when I blow the stack, run out of
memory or divide by zero. This information would enable me to measure
the size of my space (memory) and the limitations of the physics
underlying my world.
I would test the speed of each operation in my world, and find out that
some operations such as "MOVE" and "TEST IF ZERO" are relatively more
or less time consuming . From this information, I would formulate
theories on how these operations may be implemented at a higher level
and what kind of mechanism or architecture may be responsible for such
timing relationships. I would watch for white rabbits, which in my world
may be unexplained events such as "INPUTS" and try to understand what
these "INPUTS" imply and by what mechanism they may be generated.
I would generate certain events, behavior (OUTPUTS) and observe how the
INPUTS are affected by my OUTPUTS. From this information I may infer
that there is an intelligence lurking beyond my world and I may even
deduce that its name is Saibal Mitra.
George
