The Book Algorithms to Live By describes some research which found that it is not chronological age, but rather one's position on the explore/exploit curve that drives the apparent conservatism of old age. So a 70 year old who expects to live 50 more years might be considerably more novelty-seeking than a 70 year old who expects to live 20 more years, because the first has a longer interval within which to exploit the benefits of any useful finds.