I'm calling the blog "Subjective Functions"; machine learning people spend a lot of time working with objective functions, the functions we are trying optimize. One of the fun parts of machine learning is that these functions like $$\sum{(y-\hat{y})^2}$$, are influenced by our choice of models, parameters, and even what we are measuring. This means these objective functions, which seem absolute and impersonal, are actually reflections of our subjective ideas and beliefs.