Anti-inductivity is a limit case
In the past, anti-inductivity has been talked about as something a system can or cannot embody. “Markets are anti-inductive”; “authenticity is anti-inductive”; “computer security is anti-inductive.”
But like many abstract descriptors, including “cooperative” and “adversarial,” anti-inductivity is best thought of as an ideal or limit case, with real systems existing on a spectrum in between inductive and anti-inductive.
Our task, then, is to identify and list those compositional attributes that create the conditions for a system to be more or less inductive or anti-inductive.