The challenge in computational learning theory is to understand the capabilities and limitations of learning algorithms.
The problem with quantitative easing is that it works in practice but it doesn't work in theory.