Le 02-janv.-07, à 03:22, Stathis Papaioannou a écrit :
Bruno Marchal writes:
Le 30-déc.-06, à 07:53, Stathis Papaioannou a écrit :
> there is no contradiction in a willing slave being intelligent.
It seems to me there is already a contradiction with the notion of
I would say a willing slave is just what we call a worker.
Or something related to sexual imagination ...
But a "real" slave is, I would say by definition, not willing to be
OK, a fair point. Do you agree that if we built a machine that would
happily obey our every command, even if it lead to its own
destruction, that would (a) not be incompatible with intelligence, and
(b) not cruel?
Hmmm.... It will depend how "we built" the machine. If the machine is
"universal-oriented" enough, through its computatbility, provability
and inferrability abilities, I can imagine a "cruelty" threshold,
although it would be non verifiable. This leads to difficult questions.
For in order to be cruel we would have to build a machine that wanted
to be free and was afraid of dying, and then threaten it with slavery
For the same reason it is impossible to build a *normative* theory of
ethics, I think we cannot program high level virtue. We cannot program
it in machine nor in human. So we cannot program a machine "wanting to
be free" or "afraid of dying". I think quite plausible that such "high
level virtue" could develop themselves relatively to some universal
goal (like "help yourself") through long computational histories.
In particular I think that we should distinguish competence and
intelligence. Competence in a field (even a universal one) can be
defined and locally tested, but intelligence is a concept similar to
consciousness, it can be a byproduct of program + history, yet remains
beyond any theory.
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To post to this group, send email to firstname.lastname@example.org
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at