----- Original Message ----
From: Matt Mahoney <[EMAIL PROTECTED]>
Your question answering machine is algorithmically complex. A smaller
program could describe a procedure for answering the questions, and in
that case it could answer questions not in the original set of 10000.
Here is another example:
3 => 9
7 => 49
8 => 64
12 => 144
2 => 4
6 => ?
You could write a program that stores the first 5 training examples in a
table, or you could find a smaller program that computes the output as a
mathematical function of the input. When you test your programs with "6
=> ?" which program would give you the answer you expect? Which would you
say "understands" the training set?
You can take the position that a machine can never "understand" anything
the way that a human could. I don't care. Call it something else if you
want, like "AI".
-- Matt Mahoney, [EMAIL PROTECTED]
-------------------------------------------
First of all, just to make sure you understand me, when I said that the
'generator=prediction & compression=understanding' theory and the is not worth
spending much time on, I did not mean that I thought your ideas were not worth
spending time on. I was just criticizing that one idea, not all of your ideas.
Secondly, I have already considered the example that you supplied me with. I
tried to explain that I have already discussed these ideas in another group.
And I have never taken the position that a machine can never "understand"
anything. That is either a straw man argument or it shows that you did not
understand what it was that I did say in my last message. Or maybe you did not
even bother to read my last message very carefully before you fired off your
resply. I don't see any other way of explanation.
I do think a truly intelligent computer program would be algorithmically
complex. Evidently that would be a difference in our opinions if you have
completely accepted the position that you are advocating. I also think that
generalization, generalization-like relations or generative procedures are
necessary to produce AI. I think that is a similarity in our positions.
You can try to find the fundamentals of intelligence, that is of algorithmic
intelligence, but that does not mean that you will be able to produce
intelligence before you find a theory that is complex enough to explain how
artificial intelligence can be produced. That is another weakness of
compression=understanding theory.
Jim Bromer
-------------------------------------------
agi
Archives: http://www.listbox.com/member/archive/303/=now
RSS Feed: http://www.listbox.com/member/archive/rss/303/
Modify Your Subscription:
http://www.listbox.com/member/?member_id=8660244&id_secret=103754539-40ed26
Powered by Listbox: http://www.listbox.com
