Build an illustrated “map” of the AI safety territory - philosophical, technical and economic considerations