Solvent comments on Proposal for “Open Problems in Friendly AI”