Re: Good AI?

From: Heartland (
Date: Thu Dec 22 2005 - 09:13:09 MST

> On Thu, 2005-12-22 at 02:41 -0500, Heartland wrote:
>> Would it make more sense to replace CEV with the concept of goodness?
>> It
>> seems to me that implementing an invariant of goodness for an initial
>> dynamic should encompass all of the goals of Friendliness while such
>> an
>> initial dynamic system should be more tractable than building it with
>> CEV
>> (emphasis on the uncertainty of "seems" and "should").

> OK, I renamed the "Friendliness" variable in my code to "Goodness."

Not really. It might be an alternative "path" to get to Friendliness by
replacing CEV.

> How does this help?

By shifting the burden of designing a Friendly initial dynamic further to
AI, this might offer a simpler path to get to Friendliness while giving
humans less chance to screw up complex things like CEV. I'm not against CEV,
but there might be a better way to accomplish CEV goals with a different


This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:54 MDT