Ultimate question of Intelligence was Re: Is the Curt net a kind of decision tree?
- From: "Will" <wil.pearson@xxxxxxxxx>
- Date: 30 Jun 2006 03:43:34 -0700
Curt Welch wrote:
The ultimate problem that
all intelligent machines must solve is the answer to the question "what
should I do now"? Using huge decision trees to answer that question seems
reasonable to me. That's basically what these networks are.
I depends what you mean by "do". If you mean what should I output next,
then I would disagree. If do means what processing should I do now,
then you would find me in agreement. The brains processing consumes 20%
of the energy a human uses. It also has neurons that couple with the
brains vascular system and the neural coding is thought to be energy
efficient.
All these things to me suggest that the question of what processing to
do is important to human brains and human intelligence. It also means
that formalisms such as AIXI are not dealing with the same question of
intelligence as I am.
Will Pearson
p.s. I don't have time to get back to the numerous loose ends of
threads I have scattered all over the place, I apologise. I should be
learning VHDL! I can't design a system that can monitor the energy
usage of its sub systems..... but it would be better than trying to
implement my system in serial and it makes me have to deal with
parallel systems properly, so a step closer.
.
- Follow-Ups:
- References:
- Is the Curt net a kind of decision tree?
- From: JGCASEY
- Re: Is the Curt net a kind of decision tree?
- From: Curt Welch
- Is the Curt net a kind of decision tree?
- Prev by Date: Re: Is the Curt net a kind of decision tree?
- Next by Date: Re: smallest AI
- Previous by thread: Re: Is the Curt net a kind of decision tree?
- Next by thread: Re: Ultimate question of Intelligence was Re: Is the Curt net a kind of decision tree?
- Index(es):
Relevant Pages
|
|