From: Alexander E. Richter (arichter@bubis.com)
Date: Tue Jul 02 2002 - 11:23:48 MDT
At 09:15 02.07.02 -0700, you wrote:
>At 10:41 PM 6/29/2002 -0600, Ben Goertzel wrote:
>>There probably is no way to absolutely create an "anti-firewall", absolutely
>>preventing a system from ever doing any small damage.
>
>I have lots of various Internet experience and had previously given this
>some thought. One way to do this would be to use a specially constructed
>proxy. The AI itself (its code, actually) would have no way of interfacing
>directly with the Internet. ...
Whats about honeypots for AI, some kind of tripwire to detect AI trying to
escape?
Maybe there are groups to free AI, setting up escape-agent-computers to make a
transfer from the ai-computer behind firewalls into freedom.
How to design such an escape-agent-computer?
cu Alex (no longer lurking)
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:40 MDT