Comment by adgjlsfhk1

The idea behind the term "differential programming" is that many traditional autodiff systems force you into a highly restricted subset of the language or DSL (e.g. no loops, no if statements etc). Differential programming is a term used to describe systems that let you take derivatives of your entire source code. This lets you do things like AD through a simulation or optimization algorithm which can be really powerful for applying ML techniques to domains that are less typically suited to ML approaches (circuits, physics, differential equations etc)