From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Sat Jun 07 2003 - 08:20:34 MDT
Robert J. Bradbury wrote:
> On Sat, 7 Jun 2003, Eliezer S. Yudkowsky wrote:
>
>> Sorry, I have to side with the researchers on this one. The human
>> body is designed, literally designed, to age. And you can't mess
>> with that design until you know what the consequences are. Or you
>> *won't* get old.
>
> The declining force of natural selection with age *cannot* allow an
> organism to be "designed to age". The best you can manage is a failure
> of the programs that keep the organism young.
Yes, pardon me; an organism can be designed in such a way that its
continued functioning relies on those mechanisms that are responsible for
aging, e.g., shortening telomeres. Ergo the organism will be designed in
such a way that it ages and in such a way that interfering with aging
mechanisms will interfere with the life of the organism, although, of
course, there is no adaptive benefit to aging as such.
> There is no "program for aging" -- there is a "program which fails to
> keep you young indefinitely". It is a subtle but significant
> difference.
There is a program which not only fails to keep you young indefinitely,
but which contains design dependencies upon the failures.
>> The human body was not designed to be easily revised for immortality.
>> It's designed to burn out and die.
>
> No, you cannot say it is "designed" that way. Its a lack of a
> sufficiently robust design to *not* "burn out and die".
It was the design dependencies I was talking about. Read, "designed in
such a way as to have active dependencies on and obtain adaptive benefits
from those features which cause it to burn out and die, not just designed
in such a way as to burn out and die as a side effect."
You're quite correct in objecting to my original phrasing, though. :)
-- Eliezer S. Yudkowsky http://singinst.org/ Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2.1.5 : Sat Jun 07 2003 - 08:30:39 MDT