prins (02/02/83)
Prem Devanbu has appealed to the essentially private nature of consciousness
to defend his notion that consciousness can not be created within a machine.
I think the crux of the argument is that consciousness is unverifiable since
it is purely subjective.
The inability to prove the presence of consciousness does not preclude its
existence. Every few seconds, by some accounts, a proto-conscious being is
evicted from a womb somewhere. These little toddlers grow up to be all too
conscious brats. Or so we suspect, since we can't be sure.
If Prem is willing to attribute consciousness to his friends (or perhaps only
some of them), then why not a machine? How did his friends get so endowed,
consciousness being non-communicable and all that? An argument on the
non-feasibility of machine consciousness must hinge on fundamental differences
between humans and machines, not on our inability to recognize consciousness.
But perhaps Prem is a purist. Since no person or thing can be proven to be
conscious, none of them are. Then Prem, the only "I", somehow saw fit to
engage in a discussion barring consciousness from machines. Who was he trying
to convince? Just a bunch of machines without consciousness, plodding through
netnews.
/jan prins {vax135,decvax,ihnss}!cornell!prins [uucp]
prins@cornell [ARPA]turner (02/07/83)
#R:cornell:-396500:ucbesvax:1800001:000:1631 ucbesvax!turner Feb 6 20:23:00 1983 Well, but I think that Prem has a point here. First, however, we should take note of the caveat in (1) - "by mathematical or linguistic means". By which, of course, he means to imply that these are the only forms of communication to machines (intelligent or not.) A debatable point, perhaps. Where I think he's right is on the means by which any existing consciousness is communicated: from the environment (including the body) to the potential communicant, via sensory organs. The newborn-to-brat trans- formation need not take place only through communication from other intelligent beings, though the evidence (e.g., feral/isolated children) suggests that this is enormously helpful. This environment, and the extent to which our perceptions of it are shared, is what leads us to believe in each other's consciousness. This belief may be an act of faith. So what? By the way, nothing I've said here is terribly original. Most of it derives from Hubert Dreyfus's [gaak-coff-spew] critique of the AI community. PLEASE READ HIS BOOK, if you haven't already. If you have, read it again. Even AI people have made concessions in this direction (which Hubie himself only too gleefully points out.) Dreyfus's main point about AI is, I think, that not only are AI people reinventing philosophy, but they're doing it a SLOWER rate than the philosophers of over a century ago, since they have to periodically stop and write huge programs that prove very little. So read Dreyfus, read some philosophy, and then come back to your Pop-Logicians and see how wonderfully modern they sound. Sigh. Flame Off. Michael Turner