If we managed to create computers that had consciousness, how would we know?
Asked by
Mariah (
25883
)
April 1st, 2011
Someday it is likely that we will create computers that are capable of responding to any given input in a manner that is indistinguishable from a human response. But, does this make them conscious? I don’t know what exactly makes the distinction, but a computer that has been programmed to have realistic responses feels somehow fundamentally different from a conscious human.
We can’t even really prove that other people are conscious; being unable to get out of our own heads, we basically just assume that other people are conscious like we are. So if we created conscious computers, how would we know it?
Or, is the computer described above conscious by definition? Is the ability to have the “proper” output for any given input the criteria for consciousness? Or is there something indefinably different about human consciousness?
Observing members:
0
Composing members:
0
11 Answers
The computer would have to refer to itself as I. It would be self-aware of its existence.
@marinelife Ooh, interesting idea. But, you can program a computer to use the word I, without it really knowing the meaning of the word. How does one distinguish whether the computer “understands” the meaning of the word?
@Mariah It would have to be from the context of its communication. If it worried about its future, if it showed concern over being unplugged.
@uberbatman Yeah, the Turing test is I guess what I referencing by “responding to any given input in a manner that is indistinguishable from a human response.” So that’s the criteria for consciousness then? I don’t really have an argument because I can’t explain rationally why proper responses don’t seem to encapsulate consciousness for me.
@marinelife I see what you mean. If somehow we created AI that did not just respond with preprogrammed phrases, and yet used the word “I” and had a sense of self. That would convince me of consciousness.
How do I know that any of you have consciousness?
The answer is, I do not. For all I know, humans could all be automatons only appearing to be sapient.
It is the same with a computer. It may just appear to be sapient.
And appearance of sapience would be sufficient for me to consider the computer sapient.
I apply the same standard to humans, and certain animals.
I feel it would not be right to apply a more rigorous standard to computers than we apply to ourselves.
@ragingloli Right! I don’t imagine we could really know, in the same way we can’t prove past our assumption that any other person is conscious. GA.
While not exactly on topic, the question is making me think of electric sheep
i lurve electric sheep
“Try to count electric sheep, Sweet dream wishes you can keep, How I hate the night”
when it says no…without being programmed to.
It might at some point fight us when we try to turn it off.
Answer this question 
Composing members:
0
