On 5/25/2015 5:54 PM, Bruce Kellett wrote:
Bruno Marchal wrote:
On 25 May 2015, at 08:58, Bruce Kellett wrote:
Part of my problem is that the UD does not execute any actual program sequentially:
after each step in a program it executes the next step of the next program and so on,
until it reaches the first step of some program, at which point it loops back to the
start.
The UD does execute sequentially *each* specific program, but the UD adds delays, due
to its dovetailing duties, which we already know (step 2) that it does not change the
first person experience of the entiuty supported by that execution.
So if the conscious moment is evinced by a logical sequence of steps by the
dovetailer, it does not correspond to any particular program, but a rather arbitrary
assortment of steps from many programs.
?
Each execution of the programs is well individuated in the UD*. You can descrbied them
by sequences
phi_i^k(j) k = 0, 1, 2, ... (with i and j fixed).
But each step of the dovetailer is just a single application of the axioms of the Turing
machine in question: one of the
Kxy gives x,
Sxyz gives xz(yz),
for example. These single steps are all that there is in the dovetailer. But such steps
lack a context -- they make no sense on their own. You could simply claim that the two
basic steps are all that is needed -- consciousness self-assembles by taking as many of
these in whatever order is needed.
If the next step of program phi_i is some 10^50 or so dovetailer steps away, the only
thing that could possibly link these is the program phi_i itself -- the actual execution
of the steps is entirely secondary. In which case, one would say that consciousness
resides in the program phi_i itself - execution on the dovetailer is not required. I do
not think you would want to go down this path, so you need something to give each step a
context, something to link the separate steps that are required for consciousness.
The teleportation arguments of Steps 1-7 are insufficient for this, since in that
argument you are teleporting complete conscious entities, not just single steps of the
underlying program.
Of course, given that all programs are executed, this sequence of steps does
correspond to some program, somewhere, but not necessarily any of the ones partially
executed for generating that conscious moment.
Yes. So what?
I was trying to give you a way out of the problems that I have raised above. If you
don't see that the sequential steps of the actual dovetailer program give the required
connectivity, then what does? You did, some time ago, claim that the dovetailer steps
gave an effective time parameter for the system. But even that requires a contextual
link between the steps -- something that would be given by the underlying stepping --
which is not the stepping of each individual program phi_i.
I think what it boils down to is that steps in phi_{i}, where {i} is a set indexing
programs supporting a particular consciousness, must be linked by representing
consciousness of the same thing, the same thought. But I think that requires some outside
reference whereby they can be about the same thing. So it is not enough to just link the
phi_{i} of the single consciousness, they must also be linked to an environment. I think
this part of what Pierz is saying. He says the linkage cannot merge different physics, so
effectively the thread of computations instantiating Bruce's consciousness imply the
computation of a whole world (with physics) for Bruce's consciousness to exist in.
My apologies if I'm mistaking your or Pierz's ideas.
Brent
--
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.
