From: Cove Stuart (firstname.lastname@example.org)
Date: Tue May 29 2001 - 14:31:38 BST
Cove Stuart on Kyriacou Elias
RE: Dennett: Making Conscious Robots
>>> It is unlikely, in my opinion, that anyone will ever make a robot that
>>> is conscious in just the way we human beings are.
>> I agree with this statement, but does this really matter? If we are
>> using Turing's criterion as a guide then surely not. As long as the
>> function of the robot is indistinguishable from that of its human
>> counterpart, can we simply assume this is good enough to claim it is
>You cannot assume that it is good enough to claim it is conscious because
>a robot that is indistinguishable from its human counterpart is still a
>system that is defined by a set of rules, thus prohibiting it from ever
>truly being conscious, since a conscious mind can be considered to be
>caotic in its actions and decisions.
If the robot could pass TT3, then the 'rules' that governed its
behaviour would be identical to the ones that govern its human
counterpart. This is the point of Turing indistinguishability, if this
performance measure is not enough of a criterion for consciousness what
are we going to use? My point was that although Dennett may be right
in assuming that the 'brand' of consciousness afforded by a synthetic
human would not be a precise match (TT4), the performance criterion
should be enough to justify the presence of more than computation.
>> This view appears to me to be pseudo-science. Dennett correctly argues
>> that if supernatural forces had been accounted for explainable, causal
>> principles, scientific knowledge would not be as sophisticated as it is.
>I do not neccessarily agree with this view, because there is no evidence
>to suggest that both the supernatural and scientific world cannot exist.
This is very true, and I agree that both the explainable and the
supernatural have to coexist. But the supernatural is not defined by
causal principles, and it is these causally defined mechanisms of the
universe that we explain with science and engineering. From a reverse
engineering perspective, the mind may be, but probably not completely,
causal. This means the TT3 performance criterion is the best we have to
determine if our synthesis of the mind is any good.
>My opinion on MIT's Cog Robot is that it will never truly be a conscious
>thing, despite all the expensive accessories that are attached to it to
>make it resemble as closely as possible human behaviour.
I absolutely agree, but this may not be a result of the approach.
Attempting to solve problems in an engineering sense is usually very
enlightening, regarding the underlying scientific principles. The
'expensive accessories' may well be fundamental to the construction of
a conscious robot, however, in this case the sum of their parts does
not add up to something we can subject to the TT3, as it doesn't
possess even a tenth of generic human ability, and more importantly it
is not truly responsible for the dynamics of its internal states.
This archive was generated by hypermail 2.1.4 : Tue Sep 24 2002 - 18:37:31 BST