And it predicted that the most likely continuation for "trying to make" should be "conversation". Not "money", not "breakfast", not "it home", but "conversation

And it predicted that the most likely continuation for "trying to make" should be "conversation". Not "money", not "breakfast", not "it home", but "conversation". What kind of word-counting statistics on a modern english corpus would give you that outcome?