[deleted] comments on My new paper: Concept learning for safe autonomous AI