Someone has remarked that 'An ideal math talk should have one proof and one joke and they should not be the same'.