Information processing in generalised probabilistic theories II
The framework of Generalised Probabilistic Theories (GPTs) provides a unified way of understanding quantum and classical theory as well as essentially any other conceivable theory of physics. The information processing capabilities of these generalised theories has been a subject of study since the paper "Information processing in generalized probabilistic theories" [Barrett 2005]. An important tool for the study of quantum information is semi-definite programming, this has a generalisation known as conic programming which is almost exactly analogous to the generalisation of quantum theory to GPTs. We explore this analogy showing that it allows for results from quantum information to be generalised, for example, we study the impossibility of bit commitment and Wiesner's quantum money scheme in GPTs.