On 8/14/2014 5:50 PM, Russell Standish wrote:
On Fri, Aug 15, 2014 at 12:09:27PM +1200, LizR wrote:
On 15 August 2014 09:29, meekerdb <[email protected]> wrote:
On 8/14/2014 11:40 AM, Platonist Guitar Cowboy wrote:
Then it'd be no problem for you guys to clearly spell out what that
environment is.
Yes, that's a problem. The MGA considers a computational sequence that
produces some conscious thought. I think that in order for the
computational sequence to have meaning it must refer to some context in
which decision or action is possible. That's what makes it about something
and not just a sequence of events. I initially thought of it in terms of
the extra states that had to be available for counterfactual correctness in
response to an external environment, e.g. seeing something, having a K_40
atom decay in your brain. But now I've think the necessity of reference is
different than counterfactual correctness. For example if you had a
recording of the computations of an autonomous Mars Rover they wouldn't
really constitute a computation because the recording would not have the
possibility of branching in response to inputs. And the inputs wouldn't
necessarily be external, at a different state of the Rover's learning the
same sequence might have triggered a different association from memory. So
the referents are not necessarily just external, they include all of memory
as well.
Given that comp assumes consciousness supervenes on classical computation,
it's still hard for me to imagine what the difference is that
counterfactuals or meaning supply. That is, a classical computation (as
opposed to a quantum one...perhaps???) is a well-defined set of steps, and
if you re-run them in the MGA they're identical. There may be no
possibility of reacting differently to different inputs, but I can't see
what difference - i.e. what real, physical, engineering (etc) type
difference that makes. If consciousness is digitally emulable, then it can
be replayed, and whatever "counterfactuals" and "meanings" that the
consciousness may attach to its internal states or (replayed) inputs will
be repeated.
So in a nutshell I can't see how, assuming consciousness supervenes on
physical computation, that "being about something" or having "meaning" or
"needing counterfactual correctness" -- or needing a real environment, for
that matter, as opposed to identically repeated inputs -- can make any
difference to whether the UTM in question is conscious. Because a system
that interacts with an environment and one that replays that interaction
exactly are, or can in theory be made, physically identical.
What am I missing?
The consequence of assuming that counterfactuals make no difference in
your supervenience thesis is that it implies consciousness supervenes
on a recording. I constantly stumbled over this point too, as it is not
adequately spelled out in typical formulations of the computational
supervenience thesis.
That does seem strange, but I don't know that it strikes me as *absurd*. Isn't it clearer
that a recording is not a computation? And so if consciousness supervened on a recording
it would prove that consciousness did not require computation?
Brent
For some, that is a bridge too far. Maybe you could try following
Bruno's "stroboscope argument" to see if that persuades. (Not sure if
there's an English language version about, though).
Where Brent and I differ from the usual interpretation is that we
don't think counterfactuals are irrelevant to the physical
supervenience thesis either. That is because the physical world is
fundamentally quantum, and for all intents and purposes acts as a
Multiverse, so the counterfactuals are just as physically real as the factuals
But the MGA does show a contradiction between the physical and
computational supervenience thesis in a classical, low resource (aka
non-robust) physical reality, and this suffices to handle the
non-robust case of the UDA (aka step 8).
Cheers
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
