There is a very good reason to think that: the press always gets this wrong. So unless you either have some real knowledge in the field or you've come across Scott Aaronson's writings, you may be excused for believing that quantum computers "try all possibilities in parallel".