[ExI] survivor

spike spike66 at att.net
Sat May 14 19:48:19 UTC 2016

>... On Behalf Of Mike Dougherty
Subject: Re: [ExI] survivor

On Thu, May 12, 2016 at 12:40 PM, spike <spike66 at att.net> wrote:
>> ... (Saul Alinsky would agree 
> wholeheartedly (if you win, then even if your reasoning is flawed, it 
> becomes sound by definition after the fact (that’s magic of winning 
> (the same logic behind the old familiar saying the winners write the 
> history books (and propagate their DNA (both the genetic and memetic 
> varieties.))))))))

>...fwiw, I love the way you use parenthesis so unapologetically.  I assume this is how a math nerd applies the rules of PEMDAS to English.

>...ex: that quoted paragraph has 8 adjacent end-parens.  Your asides have asides to an exponential level.  :)


Ja.  Mike back when I was in college in the 1980s (uh... early... 1980s...) we were starting to talk about Artificial Intelligence.  Of course none of us knew anything about anything then.  I was told that to do serious AI, you needed LISP.  So I learned LISP.  We soon realized what the acronym was really for: Lotsa Idiotic Stupid Parentheses.  

I wondered, is there a better way to do function hierarchies?  But by then, it was clear enough that LISP wasn't getting us any closer anyway, it was yet another blind alley with a dead end.  

What we needed was some kind of effective feedback and self-modification loop paradigms which we did not really have then, and do not today (AI hipsters, do prove me wrong (go ahead, I can take it (I would be very pleased to be wrong on that.)))  Leave it to a mechanical engineer who studies controls, to point out to software people that AI needs some kind of self-referential control loops. 

>From Thune's class a couple years ago, I am astonished at how little some sub-areas of AI have advanced since those benighted times so tragically long ago.

All real intelligence, bio or artificial, needs some form of introspection and self-modification.  (Quote me on that if you wish (no need for attribution (plenty of us have recognize it (a long time ago.)))) 


More information about the extropy-chat mailing list