Personally, I prefer more of both. You have to actually do things - as in, develop the engineering, test things, see what works and what doesn't - to advance th

Personally, I prefer more of both. You have to actually do things - as in, develop the engineering, test things, see what works and what doesn't - to advance the state of the art. We need robots for science, humans for engineering.