A few years ago when Dubya was US president, one of his handlers, Cheyney or Rumsfeld or somebody caused much mirth when he talked of the things we know and the things we don’t know. The latter he divided into the things we know we don’t know and the things we don’t know we don’t know.
I laughed too but for different reasons perhaps. I believed he had missed the most important part, the things we know but don’t know we know.
Just as I was amused by those ramblings I am amused by the blind faith so many people at Medium have in ‘big data’ and ‘machine learning’ as if these people are unaware that they were once infants, and consequently believe we learn everything from formal teaching and study in an academic environment.
We learn the most important stuff by osmisis, that is we absorb information from the world around us, prinicipally language. None of us would get far in any endeavour without language, but nobody feeds big data to our algorithms. Parents and grandparents talk to us, play alphabet and number games, recite rhymes and so on but the speed with which we learn language and develop the skills to express ourselves once we have uttered our first words.
Nobody knows how the process works so nobody can ever hope to program a computer to learn in this way. Thus those who talk of data science and artificial intelligence are deluded. All their SQL routines are doing is processing data. Data science is exactly the same as what we were doing when I wrote my first program 50ish years ago, we called it data processing.
So don’t rule out intuition (after all its not that long since silicon valley was hyping is script kiddie products as “intuitive software”,) experience, common sense. Data science is just this years buzzphrase, and data processing machines got a lot smaller and a lot faster but still work in exactly the same way as when my first program crashed all those years ago.