The Unreasonable Effectiveness of Data: Opportunities for Actuarial Science

Renato Assunção

Eugene Wigner published “The Unreasonable Eﬀectiveness of Mathematics in the Natural Sciences” in 1960. he expressed his amazement that highly complex physical phenomena can be described, understood, explained, or predicted with relatively simple mathematical concepts and models. More recently, economists were discussing either this mathematical effectiveness is also present in economics. In 2009, three Google researchers made the point that we should stop acting as if our sole goal is to create extremely elegant and mathematical theories, and instead embrace complexity and make use of the unreasonable effectiveness of data. Ten years later, we see today the enormous success of this data-driven strategy in our daily lives, from the automatic translation to autonomous cars. How much of this success and research paradigm can be transported to actuarial science? How much should we embrace this view in actuarial science?