0 / 60 seg.

Once an algorithm (or a human designer) has found a problem-solving strategy that seems to work satisfactorily, should it concentrate on making the best use of that strategy, or should it search for others?