Abstract I contrast the “data fitting” vs “data interpreting” approaches to data science along three dimensions: Expediency, Transparency, and Explainability. “Data fitting” is driven by the faith that the secret to rational decisions lies in the data itself. In contrast, the...