He does write about this in one of his books (although I've forgotten which one). He tells a story about arguing with a colleague and coming to the conclusion that "trivial" in the context of mathematics is often used to simply mean "it is proven" not that it is necessarily trivial to arrive that the solution.