The 2 critical points are 1 and 20:
1. ...more data has been created in the past two years than in the entire previous history of the human race.
20. ...At the moment less than 0.5% of all data is ever analysed and used...
Data is NOT information. What we are creating is an enormous amount of noise with almost no signal.
I shared my production and analysis stack architectures 'madness' a year and a half or so ago... one little webdev can play with a lot of content and a lot of data with the right 'stuff'. However, I refuse to simply collect because I can - if I'm not planning to use it I don't (intentionally) gather it; currently I'm analysing 99.999% of new data collected and using ~95% of it directly.
Note: 0.001% of a TB is still a large number: ~10 MB.
I have seen the data I've collected double every 10 months the past few years. This year it doubled in eight.
The numbers involved in the various fingerprinting and tracking, cross associating, and what-have-you are mind boggling. Now with contextual delivery and HTTP/2 push it is not only being collected and analysed but called for input into personalised delivery as well. So what once was simply input for reports is become data transformed into information called into use on initial and subsequent connections.
The stock market has automatic trading...webdev now has automatic contextual serving....data jet propulsion.