"Michael M. Butler" <butler@comp-lib.org> posted:
> The working definition I use: _A_ singularity is when everything
> seems to be happening at once; the rate of change becomes so great
> as to be incalculable.
To me there are two drastically different important meanings
to "singularity".
The first, will happen "just over the the horizon" of the
immediate future. While the second might never happen, and if it
does, it will effectively be the end of real time.
The second one, if it ever occurs at all, will likely be
something like Tippler's omega point during which everything and
everyone becomes truly omnipotent, omniscient, all good, all qualia
experiencing... In order for this second one to occur our progress
must be "asymptotic" or such that the slope of our progress becomes
infinite before some fixed real point in time.
If our progress is anything less than this, even if it is
fantastically "exponential" where even the 3rd, 4th... (or any finite)
order derivatives (or the slope of the slope of the slope...) are
forever increasing we will never really reach the truly infinite. We
will eventually surpass any finite measure of power, knowledge,
goodness..., and rate at which we achieve such, but we will never
become truly infinite. Remember, there is a HUGE (read infinite)
difference between any paltry finite being, no matter how fast they
are progressing, and the infinite.
Consider just one of the fantastic finite tasks (any finite
task is infinitely trivial compared to the actual infinite) which must
be surpassed and become easy for us all, long before we can become
truly infinite or achieve such a true singularity. We must all be
able to simulate and calculate all possible histories of the earth and
discover which of the near infinitely many is the one true history.
In other words, we must have the ability to perfectly remember and be
simultaneously aware of the location of all atoms that have ever been
within the biosphere of the earth for it's entire history.... Of
course such would necessarily give us the ability to resurrect every
animal, including those that are not cryonically preserved, cremated
or in any way ever "information theoretically" dead. All such finite
tasks are infinitely far away from being truly infinite. And there
are infinitely many of such finite tasks that are even beyond this
distant horizon beyond which we cannot yet comprehend.
Of course, 1000 years from now, according to our currently
very limited and finite reckoning, growth, progress, and everything
will be happening "at once" or "so great as to be incalculable." But
this is only relative to our current abilities to "calculate". I
think anyone would be absurd to think we could get anywhere near the
above fantastic and still hard to make out distant horizon within the
next 1000 years.
In fact, our ability to make or comprehend or perceive such
calculations is a limiting factor which will always prevent us from
making an instantaneous jump to beyond that horizon. We must all have
the ability to calculate how to get there, which must include the
ability to at least partially perceive what it will be like, before we
can actually achieve such levels of growth.
In order for our growth to be asymptotic, or for the second or
real singularity to occur, something like the "omega" point, in which
everything, including experiential time, becomes infinitely dense,
must be achieved. And even in this scenario, growth according to
"experiential time" which is constantly accelerating according to real
time is still merely forever "exponential" according to what we
experience. The fact that we can experience 1000 years in a few
nanoseconds is what will give us the power to push that horizon which
we can still perceive out even further, giving us the intellect
required to progress at such ever increasing rates. If we can
compress our experiential time experienced during any finite real
period of time to be infinite, this will enable us to push the horizon
out beyond the "asymptote" and enable us to make the jump to the
infinite before some real time limit. But remember - there is a huge
difference between experiencing any arbitrarily large finite period of
experiential time within any arbitrarily period of real time and
experiencing an infinite amount of experiential time within that same
amount of real time.
So you see, even in Tipplerian scenarios, where we do get
there before some finite real time, according to what we experience,
we will never really get there and we will never really be able to see
beyond that ever moving horizon beyond which we can't yet ever
perceive or calculate. Things in our future will always be just more
of the same increasingly exciting stuff and will always remain within
what we can perceive and comprehend. To think that some single "AI"
will make it far beyond that horizon before it becomes friendly enough
and easy enough for it to pull the rest of us up to near it's level to
help it along seems absurd to me. Things will forever get more and
more fun, but will never really get out of our control since more and
more people will be required to help push us all along that much
faster.
So while I do like to use the term "singularity" to refer to
that period of time just beyond our horizon, which will likely remain
1000 years or so of experiential time beyond where we are now. (or
perhaps your settling on a horizon that is much more clean and only as
far away as October 11th, 2007?) Such a time being when everything
will be "so great as to be incalculable" given our current abilities
no matter how fantastically finite they become. I dislike the term
because it is quite misleading. The term inevitably causes people to
falsely fear it, as if even when we are in it, we will be lost in and
unable to perceive its infiniteness. But this is a completely false
idea since we cannot progress any faster than we can perceive and
comprehend. It's a simple logical fact that we must all be able to
comprehend some place in time and be able to progress to that point
before we can have the ability to venture beyond it or any single one
of us to get way beyond it without all of our help.
One exception to all this might be achieving something like
the ability to either go back in time, or make computers that can
calculate faster than infinitely fast. For if we achieved such, the
answers to all questions could be answered before we could even ask
them, not to mention all the other logical paradoxes of such. But
again, it'd be my bet that such will not be achieved within the next
6, 7 or even 1000 years if ever. So I'm not to worried. The
"horizon" is truly far away and not really a point. It simply becomes
more and more vague and imperceptible the further out we look. and it
is my bet that it will likely forever remain that way as things
forever get funner, faster and more exciting and more fantastic.
When you think about it this is just a twist on the debate
theists have been having for thousands of years. Is God truly
omnipotent... or isn't he? We don't know.... Yet.
Brent Allsop
This archive was generated by hypermail 2b30 : Fri Oct 12 2001 - 14:39:59 MDT