Assume you are given a file containing a trillion records. While inbuilt data structures in Python (e.g., lists, dictionaries) are efficient, why is it a bad idea to load a trillion records into a data structure and perform any necessary computations?