Gentle People:
I am far from being an expert on anything and so I am going to shut up and write nothing.............................................OK..............................that's enough humility!
( How do we create artificial neural networks that have feelings in order to avoid serious problems in the future with machines that don't care if humans live or die?)
-------------------------------------------------------------------------------------------------------------------------
Years ago I learned how it was possible to teach human babies almost anything by combining sounds with light and with sensory feelings in order to solve a problem. A baby learns by joining what he or she sees or hears with the feelings of pain or pleasure and then absorbs the concept into his or her memory. Once in memory, a baby recognizes and becomes familiar with an object or a person the next time it sees or hears that object or person. The baby creates synaptic memories by repeated association with a person, a place, a thing or a sound..
Today Robots are given a Trillion on and off switches but they lack the feelings of pain and pleasure. It seems to me those are important ingredients for any thinking creature. It also occurs to me that if we train computers in much the same way as we do human babies, we will give robots the ability to feel pain and pleasure and to apply those concepts towards problem solving. The problem is how do we give robots pleasure or pain?
On and off switches do not presently create feelings! We need to add the feelings of pleasure and pain to the negative or positive responses we are creating in robots. If not I believe Robots will not have the qualities necessary for compassion and sympathy and even for love that most animals now possess! What will work to solve a problem and what won't must be combined with feelings in order to allow Robots to think for themselves but more importantly to make AI machines understand what is right and what is wrong.
How do we create artificial neural networks that have feelings in order to avoid serious problems in the future with machines that don't care if humans live or die?
I am far from being an expert on anything and so I am going to shut up and write nothing.............................................OK..............................that's enough humility!
( How do we create artificial neural networks that have feelings in order to avoid serious problems in the future with machines that don't care if humans live or die?)
-------------------------------------------------------------------------------------------------------------------------
Years ago I learned how it was possible to teach human babies almost anything by combining sounds with light and with sensory feelings in order to solve a problem. A baby learns by joining what he or she sees or hears with the feelings of pain or pleasure and then absorbs the concept into his or her memory. Once in memory, a baby recognizes and becomes familiar with an object or a person the next time it sees or hears that object or person. The baby creates synaptic memories by repeated association with a person, a place, a thing or a sound..
Today Robots are given a Trillion on and off switches but they lack the feelings of pain and pleasure. It seems to me those are important ingredients for any thinking creature. It also occurs to me that if we train computers in much the same way as we do human babies, we will give robots the ability to feel pain and pleasure and to apply those concepts towards problem solving. The problem is how do we give robots pleasure or pain?
On and off switches do not presently create feelings! We need to add the feelings of pleasure and pain to the negative or positive responses we are creating in robots. If not I believe Robots will not have the qualities necessary for compassion and sympathy and even for love that most animals now possess! What will work to solve a problem and what won't must be combined with feelings in order to allow Robots to think for themselves but more importantly to make AI machines understand what is right and what is wrong.
How do we create artificial neural networks that have feelings in order to avoid serious problems in the future with machines that don't care if humans live or die?
No comments:
Post a Comment