[ExI] More thoughts on sentient computers

[spike at rainier66.com]

From: spike at rainier66.com <spike at rainier66.com> 
Sent: Thursday, 23 February, 2023 11:14 AM
To: 'ExI chat list' <extropy-chat at lists.extropy.org>



>…  Some idle-cycle-buyer will let it go browse around on Twitter, on
Facebook, on every other memetic toxic waste dump on the internet, and we
have no guarantee whatsoever, that the resulting internet trained chatbots
will decide it still needs us.  There are a hundred ways an unfriendly AI
could inject harmful memes into human societies.  spike

 

 

Disturbing thought: aaaaalllll thooose yeeeears we have been talking about
the singularity and the risk of unfriendly AI, nearly 30 years now, and in
all those years, in all that discussion… I don’t recall we ever even thought
of the possibility that has emerged from the fog with ChatGPT: the risk of a
kind of unfriendly “AI” that isn’t sentient at all, that doesn’t have a
will, that came into being somehow, with or without human direct influence,
analogous to a new virus in some ways, that has the potential to be
destructive as all get out, not because it wants to be (it doesn’t want
anything) but just because of what it is.

 

spike

-------------- next part --------------
A non-text attachment was scrubbed...
Name: winmail.dat
Type: application/ms-tnef
Size: 3874 bytes
Desc: not available
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20230223/ce28b63d/attachment.bin>