Comment by KMnO4

By this logic, there aren’t any existing Turing-complete language implementations, since memory is always finite.

This doesn’t seem like a helpful argument.