A Thought on Simulacra
Baudrillard thought that simulacra were simulations of a real referent. For instance, a boy emulating the behaviours of his father (that which is simulated), creating a certain behavior (a simulacrum of behaviour).
If you didn't want to operate on a representational mode of thinking, where everything is refering to something "real" in a sign system, then I have a following thought:
Simulation is a process that couples two sets of objects (Set A, Set B) with their own different elements together. A simulation-machine in this concept is a machine that tries map Set A upon Set B, and thus we can call set B the "simulacrum" in this instance. With identical sets the operation is simple, it is simple equivalency. With different sets however, the simulacrum will never have the precise elements of set A, no matter how much one might try to impose a certain order of things. So in the previous example, while child A has seen a "set A" (the father's behaviour" and processed a simulacrum (their own new behaviour), as with anything in the empirical world it is beset by difference. By comparing the simulacrum and that which is simulated, what is highlighted is everything that is not shared. However, compare and contrast what "Set B" originally was before it was mapped into a simulacrum and it will not be identical, there will be in fact a new creation.
Thus, simulation can be seen as a process of creation
edit: OH FOR FUCK SAKE DELEUZE HAS ALREADY WRITTEN ABOUT THIS: https://engl328.wordpress.com/wp-content/uploads/2012/02/deleuze-plato-and-the-simulacrum.pdf