Can you prove that there are arbitrarily many primes in arbitrarily big intervals?
This simple problem is an example of a very interesting phenomenon: if you have a large enough "universe" to consider, even randomly picked parts exhibit structured properties.
