Scientists should publish more detail about methods in general, but we're a long way from the kind of precision that would enable robots to duplicate everything

Scientists should publish more detail about methods in general, but we're a long way from the kind of precision that would enable robots to duplicate everything. A good deal of science happens with graduate students literally building their own apparatus.