Is There Any Other Bits in Type Inference That We Can't Understand?

The code that you've written today will just start running faster and you won't have to do anything to make it run faster. I would imagine that we may end up seeing some of the stuff only being enabled by default when you're doing it in advanced build or a simple build. So what I think might happen is if we get better and better at doing type inference, we might find that things are starting to reflect the static nature of stuff a little too much. We might have compiler options to turn some of it off,. Because you might start to see that things just don't behave correctly at the REPL if you go too far with all this.