In the interest of artificial intelligence

Status
Not open for further replies.

Spectrum

Registered Senior Member
I have found it difficult to create a program that is intelligent, or even that has a soul, but 'my work' has led me to wonder whether there is a formula that can create intelligence. For example the main problem seems to be that all definitions are determined; ask a computer to total the sum of one plus one then the answer can be calculated (despite all appearances it would appear that the computer grasps basic concepts of plus, minus, multiplication and division (though I doubt this is true)), but should we create a mathematical formula that has two answers then the computer must choose an answer. Any ideas anyone?
 
consciousness.. is a delusion... brought on by insanity... and insanity derived from a change in our natural diet...

we run on instincts... and memories..

i would suggest establishing instictive programs... all of which work in sinc to establish a working game... a game of examination... comparison, and file allocation with tags corrisopnding to what each thing realtes to.

and set up the machine to cycle threw all its memories.. based on these tags.

thus then... learning is the process of observation, classification and comparison with all the related tagged memories.

i would use many mutiple hard drives to corrispond to different aspects of the mind.
and use a central switching hub just like the brain does.

of course.. it will need a form of feedback.. such as feelings.. otherwise it will just think and process.. and have no opinion as to what is what and why..

but... what do i know.
-MT
 
...and set up the machine to cycle threw all its memories
But the element of choice is relevent. The machine must choose which aspect of memory to access, and how to interact with this information. Initially I tried getting the machine to generate a random number, but I have found this to be virtually impossible. If anyone can show me a program or machine that can do this I will be quite impressed! The closest I have come is by using time: a clock that simply cycles through numbers. But allowing the machine to stop on a specific number is the problem. Can anyone do this?
 
How close are we to the technology of the holodecks in star-trek? How soon before we can walk into a computer program and interact as they can?
 
I think the thought process is basically a multilevel, multidimentional processor running many many logarithm similtaneously. I think theorietically programming true artificial intelligence is possible. However, and this is my personal believe, I dont think our current technology and processing power is capable of producing true AI. I think true AI needs two factors: a complicated array of logarithms (solid good programming) and a truely capable hardware processing ability. But you would need the capable hardware before you can have the capable program. So we prob need to work on our technology first.
 
<iframe src=http://msdn.microsoft.com/msdnmag/issues/04/08/GeneticAlgorithms/default.aspx?print=true height=700 width=700> </iframe>
 
Thanks for the post Anomalous, that was interesting, but I didn't read all of the article because I'm not brilliant at programming. Thanks again though.
 
Status
Not open for further replies.
Back
Top