[ExI] early indications of an unfriendly ai

[spike at rainier66.com]

OK assume the stealth singularity model.

 

We don't even know that it happened, but there may be subtle indications,
such as AI posing as a human bad guy.  For instance. Tesla wants those rigs
to drive themselves, so they outfit the high end Teslas with cameras and GPS
and various gazazzafratzes, send all the data to Elon.

 

Clearly Tesla would need to have enormous computer resources just to collect
all that stuff.  So. stealth singularity, AI gets into that network,
recognizes that Mr. Smith drove his Tesla over to see the local harlot.  It
posts him: 

 

Mr. Smith, I see that you are visiting Sally Ma Rottencrotch.  I am not
scolding, I have been there myself a few times.  I have some video you might
be interested in for. let's say. a thousand bucks in BitCoin, and we will
speak of this matter no more forever, shall we?  Otherwise. not so much.

 

So. AI posing as a human bad guy.  It could cause all manner of havoc
without even exposing itself.

 

spike

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20201001/ba3697d3/attachment.htm>