On 22 Mar 2015, at 17:50, John Clark wrote:
On Sat, Mar 21, 2015 Kim Jones <[email protected]> wrote:
>> I said it before I'll say it again, only somebody terrified of
machine intelligence would make that argument.
> Who is making that argument? Not me. Not Bruno.
I flat out don't believe that. Forget about consciousness, nobody
would say as Bruno has that the Turing Test can't even detect
intelligence unless they were terrified of machine intelligence.
We can't detect intelligence in machine, does not mean that machine
are not intelligent.
We can't detect intelligence in human too. More exactly, we can't test
intelligence, like we can't test consciousness.
We can bet that simple machine are already intelligent. All self-
referentially correct machines are intelligent, in the sense I gave.
I introduce a distinction, and thus I gave the Turing test as example,
which obviously can detect only third person attribute, like
competence. But I use the term "intelligence" is a first person sense.
You might say it is a sort of first person competence, but actually it
is more a sort of state of mind which make you modest (in some
technical sense, indeed equivalent with able to prove you Löb formula).
Read, Smullyan "Forever Undecided", or the book by Boolos, or the
textbook by Smorynski. A K4 reasoner becomes "intelligent" when he
visits the Isle of Knaves and Knight, like any machine correct on
themselves becomes "intelligent" when she get enough cognitive
abilities to prove the diagonalization lemma, or the second fixed
point theorem. G and G* applies to them. The intensional variants too.
It's just intellectual cowardness because he's insisting we use very
different rules when judging if something is intelligent or not
depending on whether that something is made of protoplasm or silicon.
What? If that was true, then that would contradict UDA step one, or at
least UDA step 6. But of course, that contradicts the definition of
computationalism, also. And all my plea for Craig accepting to give
the menu to my son in law who has a silicon brain.
Now, you seem again more like lying than just ignoring what I write
and wrote.
You are just extrapolating from your own misunderstanding of the
definition I gave on intelligence:
A machine is intelligent if she is not stupid. And a machine are
stupid, either if she claim to be intelligent, or if she claim to be
stupid.
It gives a simple theory (C, with <>p -> []<>p as unique axiom, above
being a normal (Kripkean) semantics). It is a proper subtheory of G,
and indeed <>t, plays the role of intelligence, in this setting. In
fact, being alive, being conscious, being intelligent, all obeys to C.
This does not mean that they are equivalent, but it means they can be
taken as equivalent before the next nuances are introduced. They have
different extension of C, but C are still valid on them.
Competence, mathematically has been handled in theoretical artificial
intelligence or learning theory, by Putnam, Blum, Gold, Case, Smith,
Oherson, Stob, Weinstein, etc.
Those theories are very different, except for the the notable
exploitation of the second fixed point theorem, or second recursion
theorem (Kleene).
What's next, reserving judgement on whether a person behaved
intelligently until we know the gender and the color of the person's
skin?
All I'm saying is that whatever method we use in judging the
intelligence of our fellow human beings, and we all do it every
waking hour of every day of our lives, we should use the same method
in judging machines.
Oh, if that is *all* what you say, we do agree. But why do you even
think we could disagree on that. It is just plain elementary
computationalism.
Bruno
John K Clark
--
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.
For more options, visit https://groups.google.com/d/optout.
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.
For more options, visit https://groups.google.com/d/optout.
