When analyzing data sets from real-world applications, many classical "efficient" algorithms that run in time quadratic to the input size are considered too slow to be practical.

Design and operation of modern computer systems are increasingly influenced by external disciplines such as law and economics. In this talk, we explore two such emerging phenomena: digital privacy laws and cloud economics. I make a case for why these trends are fundamental and how they invalidate computing principles and practices that have years of precedence. Then, I demonstrate how to build systems infrastructure to manage their impact.