It doesn't waste time or memory. The key insight of this chapter is that efficiency isn't about absolute speed ("it takes 0.3 seconds") — it's about *how running time grows* as input grows.