The article discusses how to create a continuous time model from a discrete time model for open populations. The researchers explore two main methods: calibrating a continuous time model using a transition matrix from a discrete time model, and extending the theory to semi-Markov processes. They find that when the transition matrix can be embedded in a continuous time model, optimal solutions can be achieved. Additionally, they show that the process can be decomposed into terms involving the random variables of the discrete time model and time functions based on stopping times.