gRR comments on General purpose intelligence: arguing the Orthogonality thesis