I have a question about AI Safety & Ethics that may be useful for other people working with it too.
I am particularly interested in reliability and creative uses. I would like to compare approaches that people have actually used rather than just theoretical suggestions.
What trade-offs have...
