From: Robin Hanson (hanson@econ.berkeley.edu)
Date: Mon Sep 21 1998 - 17:43:46 MDT
John Clark writes:
>We all seem to agree that in a million years nearly everything would be
>so radically different that the world would be incomprehensible to us,
>the question is whether this transition could happen in a time that was
>short by human standards. I think a breakthrough in Nanotechnology
>alone could bring about such a singularity in days perhaps hours, but I
>know Robin disagrees with me and I'm not in the mood to rehash that
>argument. Eliezer has made a very strong case that an AI could also
>quickly change the world beyond recognition. I can't prove either of
>these ideas with mathematical precision but I'm pretty sure both are
>true.
To clarify, I don't necessarily disagree with your claim, if you
use a weak sense of "could" (say > 1/1000 chance). My main complaint
is that the analyses offered in favor of these claims are extremely
weak, not just by the standards of math, but of ordinary careful
arguments in other contexts. There seems to be something about social
aspects of the future that makes ordinarily careful people get sloppy.
Let me illustrate with your biology scenario.
>Maybe biology could radically transform things too, not in days but
>perhaps in 20 years or less. ... recently evidence has
>been found that a much higher genetic language must also exist. ...
>organisms as different as a fly and a mouse, who's last common ancestor
>lived 500 million years ago, nevertheless use the same high level
>language. ... This would also explain how evolution was able to evolve
>fully modern humans from our ape like ancestors so rapidly. ...
>If we understood that language then genetic engineering would be
>easy, if we want a bigger brain no need to painstakingly specify where
>every neuron should go, just write "make bigger brain" ... Ok,
>we might have to worry about a few details, like increasing the blood
>flow to the head for the larger brain, but then again maybe not, ...
>... there are reasons to suspect that once we start to learn it we'll
>learn all of it in a hurry. ...
>Once we understand that logic we may have a eureka moment and suddenly
>go from total confusion to complete enlightenment,
>... such knowledge might not be much more complex than the
>genetic code, but it would certainly be vastly more powerful and would
>change the world beyond recognition.
You take care to explain carefully what this language would be and why
we might suddenly understand it, and then when the subject of the timing
of social change, you just declare that "perhaps in 20 years or less"
it "would change the world beyond recognition." Can't you see the vast
space between granting your assumption of sudden language understanding
and having the world change beyond recognition in 20 years?
Even when you can design new humans immediately, it takes 20 years just
raise them to see how they turn out as adults. The entire world population
is not going to immediately switch to having all kids using radically
new designs, and big parts of the world will probably forbid any of their
citizens from trying them. It'll probably be at least ten years before
any radically different human design is even tried anywhere.
Consider also the huge investments we have made in infrastructure tuned
to bodies like ours. These will give us strong reasons to create new
bodies that fit in today's cars, can walk through today's doorways, that
walk at speeds similar to ordinary humans, speak using frequencies that
ordinary humans can hear, etc. Even if after 20 years 10 percent of the
population had keener eyesight, bigger brains, etc., the world would not
seem to be "beyond recognition".
Robin Hanson
hanson@econ.berkeley.edu http://hanson.berkeley.edu/
RWJF Health Policy Scholar, Sch. of Public Health 510-643-1884
140 Warren Hall, UC Berkeley, CA 94720-7360 FAX: 510-643-8614
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:49:36 MST