From: J. Andrew Rogers (andrew@ceruleansystems.com)
Date: Mon Oct 12 2009 - 16:27:04 MDT
On Oct 12, 2009, at 3:05 PM, John K Clark wrote:
> On Mon, 12 Oct 2009 "Robin Lee Powell" <rlpowell@digitalkingdom.org>
> said:
>
>> Is that seriously your argument? That a fixed goal mind *must*
>> check itself for infinite loops to function?
>
> Yes that is precisely my argument. A fixed goal mind MUST check itself
> for infinite loops, otherwise whenever the poor AI obeys a human
> command
> he is playing Russian roulette with his mind. But we've known for 75
> years that it CAN'T check itself for infinite loops, therefore the
> only
> logical conclusion is that fixed goal minds can't function. Come on
> people this isn't rocket science, it would be obvious to all if you
> didn't have infantile fantasies of being post singularity slave
> masters
> clouding your judgment.
On finite machines, many computational models will always halt. An
argument from Turing machines is silly, since your argument is
predicated on features that only exist for true Turing machines.
> A non fixed mind AI doesn't need to worry about infinite loops. It
> tries
> something for a while, makes no progress, says fuck it, and then tries
> something else.
Am I missing something major, or is this an incredibly naive analysis
where the decision to try "something else" is deus ex machina in an
otherwise vanilla algorithm?
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:04 MDT