On 30 Apr 2013, at 01:48, meekerdb wrote:
On 4/29/2013 2:18 AM, Telmo Menezes wrote:
On Tue, Apr 23, 2013 at 8:04 PM, John Clark <[email protected]>
wrote:
On Mon, Apr 22, 2013 Telmo Menezes <[email protected]> wrote:
I also believe that some isolated tribes assume everything is
conscious.
If they're right then that certainly solves the consciousness
problem and we
can move on to solving the REALLY hard problem, figuring out why
some things
behave intelligently.
I don't really understand why you insist that intelligence is a
harder
problem than consciousness.
I think John's point is that it's easy to theorize about the "hard
problem" of consciousness because the problem isn't even well defined
Well, assuming comp we know at least that a partial subproblem
consists in deriving physics from arithmetic.
This, per se, solves only one half of the mind-body problem, but the
"mind" half is 99,9% solved by the self-reference logics, and notably
the splitting truth/proof operating on some of the modal variant of
self-reference. And is the best we can do, as the machine itself can
understand why it remains a 0,1% which cannot be understood, for
logical reason.
and there's no way to test the theories because consciousness is
taken to be a first-person-only phenomenon by definition.
So you can test it. Like the WM-duplication prediction are testable,
and besides, the whole of physics remains something testable despite
it becomes purely first person (plural).
On the other hand it is pretty easy to test intelligence - we do it
all the time.
We can test competence, not intelligence.
But creating intelligence is a "hard" problem.
It could be easy. The UMs or the LUMs are born intelligent. It is our
dumbness which makes it hard to recognize them, or to listen to them
(which today still ask for some hard work and background in
mathematical logic).
The LUMs are intelligent: all they need is love :)
I guess that you define probably intelligence by competence. But the
competences are domain dependent, and there is a infinite lattices of
more competent machines, with a lot of incomparable degrees.
I think we have very solid hypothesis on
why some things behave intelligently, you explained it yourself. The
problem becomes easier if we reject meaning, and accept that
evolution
is just a mindless process of complexification.
In any case, through a modern combination of computer science,
neuroscience and biology, we know a lot about intelligence. We know
nothing about consciousness (scientifically, that is -- I know a lot
about my own consciousness).
Maybe because (scientifically) there's nothing to know. What would
constitute a solution to the "hard problem" that could be tested?
A coherent theology, including physics.
I think the best we will be able to do is to understand human
brains to the point that we can manipulate thoughts and emotions as
reported by subjects and we can make AI robots that behave like
humans and whose "character" we can design as desired. When we've
done that we'll "bet" (as Bruno would say) that we've solved the
problem.
We can bet that we are machine, but we cannot bet something solve a
problem. We must find a way to formulate the problem (like the comp
reversal does), and test the consequences provided by the theory (in
this case compare the comp-physics with what we observe. Today QM
still confirms comp, but it was close (newtonian physics contradicts
comp, as it lacks the MW under the substitution level).
Bruno
Brent
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it,
send an email to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/everything-list?hl=en
.
For more options, visit https://groups.google.com/groups/opt_out.
http://iridia.ulb.ac.be/~marchal/
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/everything-list?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
