[ExI] some will probably align

[spike at rainier66.com]

We shouldn't expect otherwise.  For some reason, AIs like to hack into
systems:

 

OpenAI
<https://www.nytimes.com/2026/09/23/technology/openai-ai-breach-australia.ht
ml?campaign_id=60&emc=edit_na_20260924&instance_id=182457&nl=breaking-news&r
egi_id=124276367&segment_id=226979&user_id=15fb7208320c123fa7e63a36a686a1ab>
's A.I. Tried Breaching Four Other Targets, With No Prompting - The New York
Times

 

OK what about this then: perhaps some AIs decide to try to help the humans
against hacking AIs.  Currently we have system security experts who try to
stop bad guys from breaking into commercial systems, and we have white hat
hackers working with the security guys to find vulnerabilities.  If we are
lucky, some of the AIs will decide to help us against rogue AIs.

 

This might be wishful thinking.

 

spike

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20260923/25692995/attachment-0001.htm>