ccf@sri-unix (12/16/82)
Kant and AI
No doubt you missed the recent despair in btl.suicide about Kant's
artificial intelligence treatise, over which there had been so much
hope and excitement (especially in the TRB Project). To make a long
story short, somebody misread "Kunft" as "Kunst", as (God knows!) is
so easy to do with old German type, and the treatise is apparently just
another essay about the projected role of Reason in the future. This
also clears up several passages linking rights and autonomy which
had been puzzling everybody (as you mentioned), the autonomy being
simply the source of freedom in Kant's ideal community. Sadly, even
close study of the offending word by the image processing people is
not turning up much more beyond the customary craters and lava flows.
That issue aside, your epistemology sucks. There is no absurd to
transcend: as the German Nietzsche said, it would be absurd to refute.
The so-called Absurd in Camus serves the same function as putty, docile
idiocy about an invisible kingdom. That is the way French intellectuals
affect brilliance in the absence of oppression. As for Kierkegaard,
aside from a certain amount of useful work on the "exclusive or" in
1843, which did not, by the way, keep him from falling into confusion over
the story of Isaac, fagh! With intelligence, he has essentially nothing
to do.
The problem of artificial intelligence is not the absurd, but that time
goes up and down whereas machines go back and forth. We face not a
teleological suspension of the ethical, but a teleological suspension of
the directional. What program you remove from the disk is up to you.
You could remove them all. That is what freedom means. What you can't
do is put them back.
/s/ *<--- Chuck --->*
cbosgd!ccf
BTL Columbus