[ExI] Wired article on AI risk

Stefano Vaj stefano.vaj at gmail.com
Thu May 24 13:49:40 UTC 2012

On 24 May 2012 02:16, Mike Dougherty <msd001 at gmail.com> wrote:

> On Wed, May 23, 2012 at 6:23 PM, Stefano Vaj <stefano.vaj at gmail.com>
> wrote:
> > On 23 May 2012 04:49, Mike Dougherty <msd001 at gmail.com> wrote:
> >> Though stupid computers, as tools, can be instructed to "act" by AGI
> >> before mere humans can counter that action.
> >
> > How do you define stupid vs intelligent?
> It's a complex algorithm.  If it quacks like a duck... and it's a
> duck... it's not that intelligent.  But if it quacks like a duck...
> and it's a wolf trying to get close enough to eat a duck... that's
> intelligent.

Cool. So, no need to be an AGI to be exceedingly performant in quack
imitation. Even though a criterium would seem more plausible that allow to
discriminate as well stupider ducks from more intelligent ducks and be
generalised to this sort of problem.

 >> j/k I'm sure nothing we build can ever be used against us even by
> accident.
> >
> > "We" who?
> The smart people on this list?  The proles on this list?  Humanity in
> general?  To be honest, I was making a counterpoint to my first
> facetious argument with a second facetious argument.

Ah, OK. :-)

But I have developed an allergy to the pronoun, especially in contexts such
as "us vs them". Nothing against the concept per se, but I want to know at
least which "team" are we referring to, to see if it is the one I
support... :-)

Stefano Vaj
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20120524/0aceabbf/attachment.html>

More information about the extropy-chat mailing list