From: Christian Szegedy (szegedy@or.uni-bonn.de)
Date: Thu Jan 31 2002 - 11:13:32 MST
Gordon Worley wrote:
> How so? If you mean simply that for every uploaded human that's one
> less AI that might have been created, true, but it only might have
> been created. You can't kill something that was never born. This is
> just wishful thinking.
>
This is true today, because it is a lot of work to create a new
intelligent being now. If there will be AIs, it will be quite trivial to
duplicate them and let them
develop into different directions. Then, there will be the simple
choice: do we want to duplicate an AI or do we upload a human? I think that
AIs will be much much more effective in using computational resources
than uploaded humans whose brains are mainly devoted to tasks
completely useless inside a computer.
If I would be an AI, I would not spend a picosecond to think about how
to convince a human to upload.In fact I would try to persuade them not
to do it,
and use the remaining space for something more "interesting".
Anyway, my original remark was meant quite sarcastic. I just wanted to
point out that the whole discussion was on a very very low level:
The problem is that the question whether to upload or not is not
interesting at all. There are a lot of other question which can be asked
regarding
uploading:
Is it uploading at all when you are uploaded, but your programm does not
get any processor time, but stored in some backup space?
Are you uploaded, if you get 1cycle processor time per year? One cycle
per femtosecond?
Will Bill Gates get a farm of supercomputers to run on, while poor
people will have to share a cheap PC to live in?
When are two uploaded being the same or different?
Etc, etc...
I don't want to discuss or answer these questions. It is hopeless. What
I mean: concepts like being death or alive won't have a meaning anymore.
Our whole today perspective will be ridiculous that time, and I don't
think that humanity will play an important role at all.
Christian Szegedy
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:37 MDT