On Mon, 26 Mar 2001, Damien Broderick wrote:
> I ask this in ignorance: how quickly could such a monster look-up
> table--listed, presumably, by a god, or someone else equal to the
> task--be looked up (magic physics not allowed in this test)?
I believe that both you and Eugene Leitl are looking at this
in practical rather than in philosophic-theoretical terms.
I think that I'm agreeing with Eugene that it would require
a much more grandiose context to achieve this than is usually
present in most SF, or even in most thought-experiments.
But as a question of pure theory, who knows what kinds of
transfer rates could be achieved in computronium? That is,
if c is still a limiting factor, it may be that far vaster
amounts of information might lie relatively close by than
we would think. (But I don't know if this would qualify as
magic physics so far as you are concerned.)
Eugene Leitl also added,
> ...computations are computations. Regardless how they're
> implemented. FPUs frequently use lookup tables for trigonometric
> functions. Does this mean FPUs don't do math?
Conway's Game of Life is a great example. It turns out that you
can speed up the Generations a lot by lookup. And it's also
known that gliders and spaceships in Life can be arranged to
do "or", "not", and "and" gates, and so forth, so that anything
a CPU or Turing Machine can do can also be emulated in Life.
Including sentience.
Yet I still believe that if we slowly shifted away from looking
up relatively small areas on a Life Board, and had one of these
incredible 10^10^10^10 size lookup tables, so that an intelligent
creature's next response needn't be calculated but only looked
up, then the creature would not be conscious during subsequent
runs, (i.e., after the initial lookup table had been created).
For a graphic example, suppose that our present Earth is only
one of 10^50 simulations that were being run to make an
enormous Eugene Leitl lookup table. When all the data from
all the 10^50 simulated Earth's were collated, every time
that you want to interact with Eugene in any way, even go
on a hike with him, all you need do at each millisecond
is lookup what he would do based upon (a) the present state
he's in and (b) the current input he's receiving from the
environment.
Sure, you'd apparently get the right behavior as output.
But the real Eugene would no longer be with us. Right now,
ask him a question, and the wheels really turn. A great
deal of logic and calculation occurs, where, significantly
a lot of information flows from state to subsequent state.
But in this lookup table, there is no actual calculation.
At each point in time, you merely form the enormous 10^30
bit long address; say 10^10 bits for the input this milli-
second, and 10^20 bits for his current state, and voila!
You fetch from the computronium his next state. But no
information flows. It's not a real calculation, I think.
Lee Corbin
This archive was generated by hypermail 2b30 : Mon May 28 2001 - 09:59:43 MDT