Myopic Bayesian Design of Experiments via Posterior Sampling and Probabilistic Programming