[ExI] malevolent machines

[spike]

>… On Behalf Of Stathis Papaioannou
Subject: Re: [ExI] malevolent machines

 

 

 

On 9 April 2014 09:14, William Flynn Wallace <foozler83 at gmail.com> wrote:

…

>…Unless there is something really important that I don't know about computers…

 

Bill there might be something really important that no one knows about computers.  Read on, sir.

 

>…it seems to me that having a machine 'wake up' like Mike in Heinlein's The Moon is a Harsh Mistress, is just absurd…

 

Plenty of people share that view, but stop short of offering any actual evidence.  We know that it hasn’t been done before, however…  read on please.

 

>…It does what it is programmed to do and cannot do anything else…

 

We already know ways to program software to learn.  Examples are learning checkers and chess programs.  If one wants to use one’s background computing cycles in this way, one can set the computer to play itself using two different chess programs for instance, and learn by seeing the results how to set parameters.  This would be an example of software that is programmed to learn.

 

>… Any other function is just some sort of mystical belief that is paradoxically held by hard scientists.  Comments?  Billw

 

Control software can be made to be self-modifying.  It does an action, watches what happens, makes corrective action with a feedback mechanism of some sort, it experiments.  Successful algorithms spawn offspring in a process analogous to biological evolution.  What part of this seems impossible?

 

I can think of no fundamental reason why a mechanical device could not in principle evolve a mind of its own.  A mind is a mechanical device, and it has a mind of its own, or is a mind of its own.  If you have some new insights on this question, do share please.

 

spike

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20140408/16af3d47/attachment.html>