Re: The skill of learning skills.

Lester Zick wrote:
Everyone has their own ideas for what an intelligence is. We need to pick the definitions that seem reasonable and strive to create something which fits the description. 

This seems to be pretty much exactly what everyone has been doing
which hasn't led to anything more than robotics. Perhaps we should
first stop to ask ourselves what the paradox of intelligence is and
solve that problem before simply defining it one way or another.

You keep mentioning the "paradox of intelligence". What exactly are you talking about? Where is the paradox?

On the contrary we would have created a definition for what we're looking for. Instead of searching for that warm and fuzzy "aha!" feeling we'd have a concrete means of judging our progress. The better our definition the better we can judge our progress. That is what I mean by tightening. 

Except plenty of people have already come up with definitions of what
they're looking for which provide a concrete means of judging progress
without creating intelligence. I thought it's what you're complaining
about. Maybe you should ask yourself whether intelligence is concrete
to begin with?

What do you mean by concrete here? Are you asking whether or not we can measure intelligence?

No amount of defining will give us the answer outright but you shouldn't expect it to. 

Of course I don't expect it to. I thought you expected it to. If you
don't expect intelligence to fall out of what you propose then it
falls under the heading robotics not artificial intelligence.

Does a compass get you to your destination? No, you still have to travel there yourself. Nothing would "fall out" of our intelligence "compass", but it is essential for getting to our destination.

Actually there is an important difference. We repleace "we know it when ..." with "the more ... the more intelligent". Do you see the distinction? 

What I see is a terminological expansion instead of a scientific
reduction. Calling something intelligent doesn't make it intelligence.

So you're categorically opposed to defining intelligence? How will you "know it when you see it" if you don't know what you're looking for?

All you're really doing is replacing one paradox - intelligence -
which you can't solve with another - skill, training, etc. - which
you think you can solve but ultimately will not solve the problem
of intelligence itself. And my question to you is how you expect to
"see" the "it" - intelligence - in whatever descriptive analogy you
substitute especially if the "it" can't be seen in the first place.


Do you actually know what a paradox is?

I make a definition of intelligence precisely because I need something (the "it") to be tangible and measurable. My skills definition might seem too abstract, but if we pick a limited domain we *could* turn it into a quantitative measure.

Its the difference between a compass that points at the north pole and a light that turns on when youre there. 

Then my question here is how do you know one is more intelligent
than the other and "the more - a light turns on when you're there -
the more intelligent"?


Your question here doesn't make sense.

-- Risujin
.

Relevant Pages

  • Re: New Robotics Website: www.amazingrobotics.com
    ... I was hoping to have everything perfect, but it seems to be a continuous work in progress, so here it is - good and bad! ... I specialize on robotics platforms built around 80/20 extrusions like those used in the original Evolution Robotics ER1 kit. ... are selling product. ...
    (comp.robotics.misc)
  • Speech & Other Delays
    ... While I am enthused with the progress that has been made the last 7 ... years with robotics, I have noticed that with Asimo & other robotic ...
    (comp.robotics.misc)