Humans are required to make decisions in a changing environment every day. We tend to use current (rather than out-dated) observations to inform our decisions. In order to study the decision making process, Veliz-Cuba et al. (2016) derived a tractable model for evidence accumulation in a changing environment. With the help of the central limit theorem, the authors were able to build a connection between statistical decision making and stochastic differential equations. In this talk, we will discuss the derivations and interpretations of their equations. Should time allow, we will consider a neural implementation of an optimal observer and show that in a sufficiently volatile environment neural populations corresponding to alternate choices interact through excitation rather than inhibition.