My fav thing about data scientists is that they’re all like “what we do is so precise! It’s math! Our conclusions are verifiable!” And then the first thing they tell when you ask about their data is how they have to make several subjective decisions about how to simplify the data so they can use it.
I'm a data librarian and even though the most advanced thing I do with data is basically "summation" I have to say that some datasets are so large that numbers have to be held as doubles and you can never precisely add them up the same way on two different machines