tailcalled comments on Core of AI projections from first principles: Attempt 1