Is Artificial Intelligence Conscious?
I truly like the study of consciousness, though it is safe to say no one really
knows what it is. Some philosophers has avoided the problem by claiming consciousness
simply doesn’t exist. It's the ultimate escape clause. However, the
"therefore, it does not exist" argument also applies to
"truth", "God", and even "reality" all quite
beyond a consensus description for at least three millennia. For each issue or
problem defying description or understanding, simply escape the problem by
claiming it doesn’t exist. Problem solved or problem avoided? Alternately, as
Daniel Dennett explains consciousness as an account of the various calculations
occurring in the brain at close to the same time. However, he goes on to say
that consciousness is so insignificant, especially compared to our exalted
notions of it, that it might as well not exist[1].
Oh, well.
Getting back to consciousness, most of us
have views (aka theories) of what it may be, but there is no consensus amongst scientists
or philosophers, though great novelists must have clear and resonant views. Our
task here is not to offer a fresh definition, but to suggest a starting point. Sometimes, before clarifying a difficult
concept, it is helpful to consider precursors or early signs of it. For consciousness, consider the following
stages.
A.
Stimulus-response.
(Lower organisms on up.)
B.
Belonging
(herds and pack predators, sociability)
C.
Can
mimic behaviors (beginning of learning)
D.
Learn
new strategies or use tools (mostly predators, first stage of innovation)
E.
Emotional
states (as in internal feelings by many animals and us)
F.
I am
or self-identity (first signs of consciousness)
G.
Self-reflectiveness
(internal self-communication)
H.
Consciousness
(humans)
I.
Creativity
(higher-order consciousness)
Artificial Intelligence (AI), seems
locked at Stage A, though with its marvelous language ability, it seems more
than just Stage A. However, with training, it learns more. (Think of the single-celled
Ameoba, with the Harvard library of knowledge and the writing skills of an English
professor.) But does it learn new strategies for its responses or how it trains?
Does it reflect on what it spews out? Does it have a self-identity? I have read some of these codes are now
writing new codes – no humans involved. Maybe this counts.
A colleague has suggested that AI
can generate conjectures and ideas at random, and then decide on their validity. Indeed, this seems to jump from Stage A to
Stage I, bypassing any need for consciousness.
Interesting thought, yes? Further still is the fact that AI is in its
very, very days, and its future is an unknown unknown.
Hoping this puts a few things into
perspective.
Comments
Post a Comment
Please Comment.