Example: A clingy agent might not only avoid breaking vases, but also stop people from breaking vases.
First time I read this, I thought it ended with “but also stop people from saving vases from being broken.”
The measure should equally penalize the creation of high-impact successors.
But less so if they have the same limitations?
Like people having a way to shut them down?
Somewhat, yeah. One would imagine the penalty should be equal to the expected impact of the successor over time.
First time I read this, I thought it ended with “but also stop people from saving vases from being broken.”
But less so if they have the same limitations?
Like people having a way to shut them down?
Somewhat, yeah. One would imagine the penalty should be equal to the expected impact of the successor over time.