[ExI] Science or Science

[William Flynn Wallace]

On Mon, Nov 12, 2018 at 7:44 PM John Clark <johnkclark at gmail.com> wrote:

> On Mon, Nov 12, 2018 at 7:49 PM William Flynn Wallace <foozler83 at gmail.com>
> wrote:
>
>
>> > *even the behavior of an extremely intelligent creature or AI will
>> show reliable patterns, and from those we will have some predictive
>> ability. *
>>
>
> Not necessarily. If you're playing chess with a brilliant chess AI you
> won't be able to predict what its next move will be, if you could you'd be
> as good at chess at it is and you're not,  but after it made its move you
> can easily see that it wasn't random and was in fact pretty damn smart. You
> can't predict what Stephen King's next novel will be but when it come out
> and you read it you can be pretty sure it was not made by a monkey
> banging on a keyboard.
>
> John K Clark
>

I disagree strongly with all of that.  What you can predict with a great
deal of accuracy is the moves the AI won't make, as they would be stupid
and self-defeating.  In fact, I think you could probably make a list of
what it will do and that will contain the next move much of the time.  I am
assuming that only a few of the moves of the AI are big surprises.

As for King, you can predict with almost certainty that he will write a
horror story, because that's what he does.  You could find out if he writes
about the same things, or if each book is different, and that will narrow
the choices he will make.

Predicting what they won't do isn't as informative as what they will do,
but it's not nothing and it's not random.

bill w


>
>
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20181113/27bd1961/attachment.html>