From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Wed Apr 21 1999 - 13:34:29 MDT
Lyle Burkhead wrote:
>
> Rabbits, in their present form, will probably cease to exist, eventually.
> So will humans. Geologists have found that more than 90% of the fossils
> that are found in the permian layer are not found afterward. Extinction
> events do happen. The most recent major one was about 65 million years
> ago, when the dinosaurs disappeared. There have been quite a few others,
> of varying severity. It looks like we are headed for another one.
Lyle, the last major extinction on this planet was only fifty thousand
years ago. The Neanderthals ceasing to exist is of infinitely greater
consequence than a few million species of fish being wiped out.
> If you think the human race is going to be wiped off the face of the earth
> suddenly, as a result of Drextech, then your logical course of action would
> be to go to Palo Alto and kill Eric and Ralph, and everyone else who is
> working on nanoscale replicators. This would (a) shock the nanotech
> community and bring research to a halt, at least temporarily, and (b)
> generate enough publicity to make everyone aware of the danger of Drextech.
> If you don't do this, then you have to admit that you don't really believe
> what you are saying.
Maybe I just confine myself to different methods, Lyle. It's like
refusing to proclaim myself a messiah. Do you really understand why I'm
not doing that, or do you just think I'm being an idealistic fool?
> In fact the emergence of a new life form isn't going to happen suddenly
Is this an article of faith? For Drexler's sake, why not?
Anyway, it didn't happen "suddenly". The power of humanity and
humanity's technology has been building for the last fifty thousand
years. All of a sudden, fifty years ago, we gained the power to blast
mountains into craters by pushing a button. Over the course of decades,
the planet itself was threatened by our rising power. Our technology
has become strong enough to tear "life as we know it" apart. I'm not
worried about the threshold suddenly being crossed between destruction
being impossible and possible. I'm worried about the threshold being
crossed from planetary destruction taking a deliberate and extremely
expensive effort, and between it happening in one laboratory because of
an accident.
> Any new form of life will face the same
> constraints.
Again: Why doesn't this prove that a submarine can't outswim a fish?
Don't they face the same constraints? Why are tanks such more efficient
killing machines than lions and tigers? Why can nuclear weapons
penetrate a turtle's shell?
> Any organism can only eat certain things, can only exist in
> certain environments, can only grow so fast, etc. Damien Sullivan says
> that diamond dust will have to find some way to protect itself from sparks,
> otherwise it will blow up like a grain silo. This had not occurred to me,
> but it's just another example of the constraints that all organisms have to
> deal with.
Of course goo has constraints. If the constraints on goo are
sufficiently weaker than the ones on giant, awkward, soft-biological
life, we're dead.
Sullivan's argument got torn to bits, by the way.
> What I call "calibration" is basically establishing inequalities, a
> fundamental tool in the mathematical sciences. It doesn't need to be
> defended.
If you are under the delusion that "calibration" as a mental tool is in
the same class as the manipulation of mathematical inequalities, I don't
see how you could. I would put it in the same class as Plato's proof of
the existence of the immortal soul, frankly. Pick the right "analogy"
or "calibration" and you can prove anything.
--
sentience@pobox.com Eliezer S. Yudkowsky
http://pobox.com/~sentience/AI_design.temp.html
http://pobox.com/~sentience/singul_arity.html
Disclaimer: Unless otherwise specified, I'm not telling you
everything I think I know.
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:03:35 MST