What is one way to avoid low-level parallelization details?
a. monads
b. literal functions
c. partially applied functions
d. parallel collections
