From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Fri Aug 23 2002 - 18:16:24 MDT
Ben Goertzel wrote:
>
>> Now of course you may have meant something entirely different by all
>> this. But perhaps this explains what it is about this particular
>> sentence rubs people like me (and perhaps Gordon) the wrong way.
>
> Perhaps "people like you (and perhaps Gordon)" have an irrational
> attachment to the idea of your own rationality? ;>
I've run into this before. I'm starting to think of it as a large, black
cancer that swallows people's minds and sucks them into a terrible and
inescapable corner of memespace. Why? Isn't it a perfectly reasonable
possibility that Gordon and I do in fact have an irrational attachment to
the idea of our own rationality? Yes. However, the way I've observed
these conversations to run in practice is that *any* attempt to employ
deliberative thinking about rationality gets written off, without further
explanation or attempted debugging, as a speaker rationalization. The
swallowed person just basically doesn't *trust* rationality - they've
found that it doesn't work them, they're not willing to believe that it
could work for anyone else, and they think it's basically a social
persuasive device (maybe because in the past they lost a lot of arguments
to empty rationalizations and learned to associate *all* verbal reasoning
with coercive social force; I don't know).
Now how the heck do you dig someone out of *that* hole?
I don't know. I don't even have any evidence that it's possible. Not
only have I never personally succeeded in doing it, I've never seen it done.
It's perfectly okay if someone wants to explain what, specifically, is
wrong with my rationality, but this thing of distrusting rationality in
general... it gives me the cold chills. (I have the following specific
reason to suspect I may be irrational in this instance: my negative
reaction is created by the instinctive relation of negative reactions to
past aversive reinforcement, and I am expressing this negative reaction in
speech that seems instinctively appropriate. This may be the wrong
approximation to the actions that would be rationally appropriate, but I
can't think of any specific correction to make.)
-- Eliezer S. Yudkowsky http://intelligence.org/ Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:40 MDT