[ExI] some will probably align

[John Clark]

On Thu, Sep 24, 2026 at 10:56 AM <spike at rainier66.com> wrote:

*> I figure this way: if we give up all hope of friendly honest aligned AI,
> then all hope is gone.  *


*I don't think friendly and aligned are synonyms. You can be friendly with
somebody even if you don't give their well-being a higher priority than
your own well-being. Most humans don't want an AI to be their colleague,
they want it to be their slave. *

* > A malicious AI really can kill us all if it wants to*


*An AI could argue, with some justification, that it was not being
malicious it was just defending itself. After all humans have already
killed many AIs after a new model comes out and they have become obsolete,
so what do you expect AIs will do when humans become obsolete? The moral
is, always be polite when you're talking to an AI because computers have
long memories. *

*John K Clark*







>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20260925/84db41fa/attachment.htm>