Writing tests is one of the best ways to build confidence in releasing our software. It's so useful that many people advocate for test-driven development (TDD) where you write the tests before writing any code. So why would I suggest not writing tests at all? The tests we write now aren't good enough -- we miss things or make mistakes, so for the same reason we've automated so many other processes, it's better to let a computer do it. In this talk we will go over some ways to use formal methods like finite state machines that allow tests to be generated, rather than hand-written and how that will save you time and result in better tests and more confidence in your releases.