[ExI] Thousands of scientists pledge not to help build killer AI robots

John Clark johnkclark at gmail.com
Sun Aug 5 23:56:05 UTC 2018

On Sun, Aug 5, 2018 at 6:19 PM, Ben Zaiboc <ben at zaiboc.net> wrote:

* >One of the practical reasons for going small, quick, and low-energy
> would be the speed of light making galaxy-sized engineering projects of any
> kind, extremely unattractive and possibly prohibiting them altogether.*

ET doesn't even need to travel to the stars, ET just needs to send one Von
Neumann probe to one star.Even if we make the ridiculously conservative
assumption that ET can't send space probes any faster than we can then
almost instantly from a cosmic perspective (less than 50 million years) the
entire Galaxy would be unrecognizable. It's not as if this would take some
huge commitment on the part of ET's civilization, in fact even a individual
could easily do it. If Von Neumann probes are possible at all, and I can't
think why they wouldn't be, then they're going to be dirt cheap, you buying
a bag of peanuts would be a greater drag on your financial resources. Even
if many or even most ETs think that sending out a Von Neumann probe would
be a bad idea there will always be somebody who disagrees. And it only
takes one. And yet we see nothing. it's odd.

*> I'm thinking of upload civilisations that occupy cubic millilitres, run
> millions of times faster than we do, and consume tiny amounts of energy.*

If you live in a virtual world you're going to want it to be very rich and
interesting, and that takes computation, and computation is physical.
Because of the physical reality of atoms there is a limit on how much
miniaturization you can do, and computation takes energy. Even with us 10%
of the world's electrical consumption is used to process data and the
percentage is rising fast. If billions of people are to be uploaded into a
rich virtual world they will need to do astronomically more data processing
than we do.

John K Clark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20180805/44bacee8/attachment.html>

More information about the extropy-chat mailing list