How should we understand the notion of an open future when we say that we would like to have an open future? In this talk I will consider several different ways of answering this question. I will argue that, contrary to what many people think, the relevant sense of an open future does not require truth-value gaps. Instead, I will argue, the relevant sense of an open future involves an ability to access different possible worlds. And I will also argue that an ability to access different possible worlds is best understood in terms of the notion of agent causation.
