I read the question raised recently by the American Statistical Society (ASA) - As a statistician, what is your super power - and after reading the numerous answers, all of them very geeky and not a single one mentioning producing yield, ROI, or added value, I could not resist to post this question on DSC, for data scientists. I could not believe that no one had any concern about delivering sustainable, scalable (automatable) added value, as opposed to beautiful charts and models.
I believe that real data scientists care more about delivering value (at least solutions implemented as black box algorithms, such as automated arbitraging), as opposed to models. My answer to ASA's question was:
Delivering measurable added value in large amounts, and getting paid accordingly (that is, very well).
I believe that this is one important point that differentiates us from some statisticians and fake data scientists. Haters will claim that people like me are just using a bag of tricks to generate revenue and quality traffic. In my case, anyone can do what I do, just check out our data science apprenticeship program to learn some of those tricks. It can be done remotely, from home, and with no investment. Yet nobody has managed to successfully compete against us: companies like Gigaom burned $40 MM in VC funding, and went bankrupt; Kaggle laid off one third of its staff (they've received $10 MM in VC funding, and are managed by data scientists). The reason is simple: you simultaneously need a modern/efficient toolkit, domain expertise, automation, and business acumen to really succeed, to really be a data scientist. It is far more important than being an expert in 50 types of linear or logistic regressions. And it can be acquired after a few years of horizontal business experience.
Below is the answers to ASA's questions. What would be your answer to the question: as a data scientist, what is your super power?
Answer provided by ASA followers: