[ExI] Fwd: ai

[Anders Sandberg]

Again, I am not talking about perfect performance. It is enough for Siri 
to recognize what kind of questions it is unlikely to give a 
satisfactory answer to (essentially just a supervised learning problem 
of mapping questions to user satisfaction) and say it. It will sometimes 
make mistakes, but that is practically OK. Note the huge gulf between 
mathematical in-principle arguments and actual computational feasibility.


On 2016-03-13 22:32, John Clark wrote:
> On Sun, Mar 13, 2016 at 2:26 PM, Anders Sandberg <anders at aleph.se 
> <mailto:anders at aleph.se>>wrote:
>
>     ​ > ​
>     In principle a learning system might even be able to learn what it
>     cannot do, avoiding wasting time 
>
>
> ​ But Turing tells us in general that can not be done. Maybe looking 
> for a example to prove that the
> Goldbach conjecture
> ​ is wrong is a waste of time and maybe it's not, ​and maybe looking 
> for a proof that
> Goldbach
> ​ is right is a waste of time and maybe it's not; and maybe Goldbach 
> is true but not provable so both things are a waste of time. T
> hat is why a AI, or any form of intelligence, needs both the ability 
> to get bored and the ability to disobey an order.​
>
> ​ John K Clark ​
>
>
>
>
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat

-- 
Anders Sandberg
Future of Humanity Institute
Oxford Martin School
Oxford University

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20160314/2cbad596/attachment.html>