The authors argued that a vast amount of data which exist with the help of technology, can benefit us by predicting the future. Sound like science fiction, eh?
Since Aristotle, human are struggling to find the cause in every thing. Then came the advent of statistics, which we select samples from population, inferred from those samples, and then generalized it to the population. The cause that happens in the samples is attributed to the general population. These process is done simply because collecting all the data would take huge amount off effort and wastes a lot of time.
In our world today, where data is well documented, taking the data from the population (N=all) is a viable option, and much more useful. From these data, we can see the correlation between phenomenon. Because the data is in huge amount, causality inferential is less relevant. By looking at the correlation, we can "predict"" the future.
I'm not totally agree with those claim. Causality and correlation study should go together, to enrich each other. So that we can make a better inference and better decision about any phenomena.
Here is Viktor Mayer-Schonberger and Kenneth Cukier talking at "Talks at Google".
No comments:
Post a Comment