From: Marc Geddes (marc_geddes@yahoo.co.nz)
Date: Sat Jul 23 2005 - 03:30:30 MDT
>Theorem: The two sets of meta-properties (Efficiency
>and Predictive power) and (Goals promoting expansion
>of complexity in goal system and goals promoting
>reduction in goal clashes with other agents) are
>correlated IF AND ONLY IF it is possible to fully
>convert the descriptions of computational functions
>into descriptions of cognitive agents and visa versa.
>Damn, now we're getting somewhere. Maybe.
>I read this as suggesting the correlation between
low-level logic
>(functions) and high-level volition (agents) is real
>only if Friendliness
>(reduced volitional clashing) is dependant on
>prediction and intelligence
>(extrapolating complexity) is dependant on
>efficiency.
>
> He earlier suggested a deeper, perhaps quantum
>computative nessecity for
>Friendliness which goes beyond deterministic
>prediction, he called it
>projection. I'd say this theorem is obvious to most
>people on the list, but
>that Geddes himself doesn't believe it. How very
>devious.
>
>
> Patrick
Nah, you've totally misinterpreted me. I'm sure you
meant well. I was simply putting forward another
formulation of the 'Objective Morality' conjecture.
I pointed out that there are two different levels of
organization for programming languages:
*Agent languages (dealing with concepts about
'agents')
*Ordinary programming languages
I was pointing out that there has to be a way to
translate between these two types of languages, and
suggested that Friendliness naturally falls out as a
logical consequence of the fact that you *can*
translate between the two levels of organzation ie
Intelligence imples Friendliness and visa versa i.e
Objective Morality.
No one is listening to me any more, so best leave the
dead horse alone O.K? Thanks man.
Just to clarify, I also don't think there's anything
mystical about intelligence or consciousness. I think
consciousness is entirely dependent on physical
proccesses and that intelligence is completely
computational. No mysticism required.
A short point on Intuition:
Whilst I made many silly wild guesses on list, I've
been totally consistent about my main 'intuitions'
ever since I first came to SL4. Main points:
*There is an objective (universal) value system
*General intelligence without qualia is impossible
*The Bayesian probability framework is seriously
incomplete
I added some another big consistent 'intuitional'
assertion a while back when I started repeatibly
saying:
*There are 7 universal knowledge domains
Just to clarify what I mean by this:
*There are 7 fundamentally different ways to
mathematically represent objective reality. Each of
these 7 different mathematical formulations provides a
different but equally valid perspective. None the
less, these 7 different mathematical view-points all
map to each other. By 'map to each other', I meant
all 7 view-points are logically equivalent....almost
;)
So far I haven't heard a single logical argument which
proves me wrong. Surely if I'm 'nuts' it should be
easy to prove me wrong.
I certainly look forward to finding out just who among
us on SL4 was 'stupid' , once the full theory of FAI
is completed.
--- THE BRAIN is wider than the sky, For, put them side by side, The one the other will include With ease, and you beside. -Emily Dickinson 'The brain is wider than the sky' http://www.bartleby.com/113/1126.html Send instant messages to your online friends http://au.messenger.yahoo.com
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:51 MDT