I think I meant the OP didn’t make very clear the distinction between “no training data” meaning “data exists, and humans can tell good from bad, but it’s hard to find/organize” vs “we don’t have a clear definition of good, and it’s hard to tell good from bad, so we honestly don’t think the data exists to find”.
Ah I see, thanks for the clarification. Personally I’m uncertain about this, and have some credence on each possibility, and may have written the OP to include both possibilities without explicitly distinguishing between them. See also #3 in this EAF comment and its followup for more of how I think about this.
I think I meant the OP didn’t make very clear the distinction between “no training data” meaning “data exists, and humans can tell good from bad, but it’s hard to find/organize” vs “we don’t have a clear definition of good, and it’s hard to tell good from bad, so we honestly don’t think the data exists to find”.
Ah I see, thanks for the clarification. Personally I’m uncertain about this, and have some credence on each possibility, and may have written the OP to include both possibilities without explicitly distinguishing between them. See also #3 in this EAF comment and its followup for more of how I think about this.