From: Eugene Leitl (Eugene.Leitl@lrz.uni-muenchen.de)
Date: Wed Mar 19 1997 - 00:26:23 MST
---------- Forwarded message ----------
Date: Wed, 19 Mar 1997 00:37:52 -0500
From: Robert I. Eachus <eachus@spectre.mitre.org>
To: nanotech@cs.rutgers.edu
Newgroups: sci.nanotech
Subject: Re: incomprehensible changes?
In a very well thought out reply Chris Phoenix said:
> Of course, this is a straw-man argument, because the people who
> predict gods are basing this not just on the progress of
> technology, but on the creation of new and immensely powerful life
> forms and thought forms. I won't go as far as bugs vs. gods, but
> I'll agree that 2-year-olds vs. adults is a reasonable
> comparison--once those new life forms become possible.
Those new life forms are possible, and they are us. Twenty years
ago, it would--and did--take me months to do the some of the things I
currently do several times per week. The web has become an extension
of my memory, and I use it as such. I may not seem godlike to you, or
to me for that matter, but as for the me of twenty years ago, he would
feel like your 2-year-old above. Not dumb, but with decades of
learning required to catch up.
Also twenty years ago the P?=NP issue was very hot. Today, we are
working on quantum computers where the answer probably not only will
be P=NP, but P=NP=O(1).*
> ...If we have as much as a month between large-scale nanotech and
> godlings, there is a possibility of making choices that will
> affect the kind of godlings we create. For example, it is worth
> thinking about the effect that food machines and utility fog might
> have on society, even if that period of history only lasts a
> month.
Nice image. I've always looked at it this way: How long will it
take for computer technology to get to the point where I can design an
Drexlerian assembler in my spare time? I figure less than 20 years.
Will the technology be there to convert than design into silicon or
whatever? Sure. If the tools are not available, designing and
building some of them will be an extra weekend project by then.
I'll give you a different problem to worry about... There are
people of all ages today who are falling off the technological
bandwagon. Those who manage to stay on the tiger (Mixed metaphores?
Why not?), are getting further and further ahead of those not on
board. Will there be a point at which no one not already on the train
can get on? Sure looks like it. That is the problem I worry about.
Not that machines will become like gods and obsolete the human
race. The direction we are approaching the future from means that
systems including people and machines will always have the edge. But
that we need to constantly work at making it possible for our children
to have a place in the future. Today's educational system is woefully
inadequate to prepare anyone for life in the eighties. It will take a
revolution (and one is going on but a lot of the educational community
is trying to stay out of it) to provide an educational system adequate
for the next century^H^H^H^H^H^H^H decade.
> Will the future simply be an amplification of the present, or will
> it be completely chaotic and unpredictable? It takes only one
> successful gray goo or engineered virus to make it
> unpredictable--but barring those, things we do today might have an
> effect on whether the first gods care about us or not, and things
> we do in the time between large-scale nanotech and large-scale
> intelligence will almost certainly have an effect.
I'm still not convinced that it is possible for the two to arrive
in that order. And at fifty-one, I may not make it into your first
gods category, but I suspect that a majority of those reading this
will.
There is a story I like to tell to put this in perspective. At my
sister's eighteenth birthday party, we were all gathered around the
television waiting for Neal Armstrong to open the Eagle's door. I had
worked on the ACG (Apollo Guidance Computer), my father had designed
and built power supplies for the radars at Cape Canaveral (before it
became Kennedy Spaceflight Center). My mother's father, Bud
Eichelburger, also there had literally learned to fly from the Wright
Brothers, and flew fighters in France in WWI. His wife, also there
knew all the original seven astronauts, and had flown with them (but
on commercial flights). That, however, is another story.
For something to do while the pressure was being pumped down in the
LEM, I started telling some of the story behind the story of the
camera that would record the first steps on the moon, and why this one
was black and white. In particular, I explained that although Apollo
8 had shown the advantages of taking a color TV camera to the moon,
that even getting this black and white camera on the LM had been
difficult, and it almost hadn't made it.
At this point the camera was turned on, and my mother read the
caption... "Live from the moon! When I was a child we said that
something was as likely as flying to the moon when we meant that it
was impossible."
Her father interrupted. "Helen, remember the night I came up to
your bedroom and said that an Army budd of mine was leaving on a
flight, and I might never have a chance to see him again? Then we
took the train up to New York to see him off? I never though
Lindburgh would make it!"
A few years later I was sitting with Bud, watching Voyager II
pictures captioned "Live from Saturn". I quoted his previous line
about Lindburgh and we all had a good laugh. Since then, whenever
someone has told me that something wouldn't happen in my lifetime, I
just laugh.
------------------------------------------------------------------------
*For those of you to whom this is so much gibberish, a translation
that any competent complexity theorist will cringe at: Twenty years
ago: If P is not equal to NP, there is a large set of problems which
we can never hope to solve. If P = NP, then those problems are
theoretically but not practically within reach. Today: We are solving
problems that twenty years ago looked to be totally beyond the pale,
and quantum computers will be here soon. If quantum computers can
solve NP-complete problems as seems probable, then any mathematical
problem with an answer that can be easily shown to be correct will be
computable in under a second.
For scale, the largest hard composite numbers, factored in days or
weeks 20 years ago, are today factored in seconds. The most recent
results in this area used thousands of MIPS years of computer time.
Quantum computers may soon solve that problem in fractions of a
second. To give you a practical example of why this is important even
ten years ago "routing" a chip or a PC board could take weeks of CPU
time, even with a dedicated engineer to help. Showing that a given
routing meets all the design rules is in NP, as is the design of a
working assembler.
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:44:17 MST