Re: Effective(?) AI Jail

From: James Rogers (
Date: Tue Jun 12 2001 - 23:04:53 MDT

On 6/12/01 1:56 PM, "" <>
> Assuming that the computer has no way of manipulating the physical
> environment within the box (which could endanger the human), it
> would seem that such a setup would be a safe way of experimenting
> with non-friendly AI.

Out of curiosity, exactly what would one do with an unfriendly AI? If it is
known to be unfriendly, then I assume you couldn't trust any information
that came out of it. Therefore, it would seem to me that the only use for
an unfriendly AI would be a code/data cadaver that you could inspect for
study. Interacting with a running version wouldn't seem to yield much

-James Rogers

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:36 MDT