Thank you for the feedback. This was a surprisingly useful line of interaction.

The first thing it did was make me remember that inferential gaps take caution at the very least to cross. Another way I failed was in not carrying out my empathetic modules of people far enough; I knew people would realize what I was after was large and vague, but then trailed off into assuming people would actually want to rattle off in some randomly chosen direction available to them. Taken on iota more of a step and I can feel how annoying such a prompt is.

And then I recalled something about A.I. safety; something along the lines of not being able to specify all the ways we don’t want an AI (genie?) to act; the nature of value or goal specification is too exclusive to approach from that direction efficiently. Reflection to see if I can be coherent about his will have to happen later.

As of this moment (2 am) it is unattractive to see if I am on to something or not. Thank you once more for the feedback. It feels like I’ve gained valuable responses.