The above chevron style chart represents the changing views on everything about data-How it is collected? to How it can be used or should be used? I find the number three "arrow" most interesting. Data was collected by data "designers" who were experts of their fields. And the purpose of collecting data was known prior to the endeavor. The paradigm shift of note is the combining of previously collected data (which might have had a completely different reason for its collection) with another data set that again could have been obtained for different purposes to create a new data collection that can serve a new set of purposes that was not even known when the two data sets were collected. This shift is significant in that it makes the process of getting and using data cheaper and faster resulting in a much quicker turnaround times.
"What would be a good example?" you ask. The well known flu season maps available on Google trends combined with the weather maps to see how severe flu in your area might be - is a good example of this changing paradigm. Do you have any to share?