I found it very hard to explain the Ramanujan-Petersson conjecture in a straightforward way.
I can only say now: think about automorphic forms as sound waves, and then the conjecture predicts that if the sound wave is decomposed into harmonics, no single harmonic is too big.
Is there any deep reason that it must be true other than citing Langlands' conjectures?
