Meanwhile machine learning is the high interest credit card of technical debt, what could go wrong? http://research.google.com/pubs/pub43146.html http://researc

Meanwhile machine learning is the high interest credit card of technical debt, what could go wrong? http://research.google.com/pubs/pub43146.html http://research.google.com/pubs/pub43146.html I entered stats through econometrics. I love econometrics. Econometrics is seriously hampered relative to other applied math fields by the inability to apply large scale casual experiments. For this reason parsimony is soooo important in modelling. Machine learning is the opposite of parsimony. Machine learning is wonderful for many applications, it lets us shrug off a first error of model selection, but it's not a panacea to the math problem and I think it actually might make it much worse.