The Art of Doing Science and Engineering: Learning to Learn

by Richard Hamming

Not that you should merely work on random things, but on small things which seem to you to have the possibility of future growth.