Warning: this is going to be another one of those posts that starts from a point of moderate easy-going bewilderment and ends up in a place of complete bewilderment and rage. Again, this debate seems to be highly artificial with all sorts of people making all sorts of claims and counter claims and setting up all manner of straw man arguments all over the place. With the amount of straw men being made and destroyed on this course, things are going to get messy!
Big data in itself seems to mean a heck of lot of different things to different people. I'm going to work from the idea that it is any kind of data that is problematic because of either its volume, its variety or its velocity in being produced, or any given combination of these three things. For me then, to put it into terms I can actually understand, "Big Data" is an adjective, not a noun. Big data is not an object; it is a way of describing an object. And it is not immovable or permanent or unchangeable. Data described as "big" today will not be described as "big" in ten years time when clever techies have developed a solution to it which is then in-built in the production of the data.
And that seems to be what the sensible, non-straw men, are arguing. That "big data" refers to data that does not have an in-built mechanism in the way it is produced to make it easy to analyse and process.
However, this is not to say that "big data" will disappear as we develop solutions because our development of technology will not stand still. The volume and speed of some data will probably continue to outpace us. Moreover the type of data we are trying to analyse will continue to diversify.
Of course Big Data is pretty much analytics and business intelligence. It just requires a different methodology because the analysis is not as straight forward. To say though that it is meaningless and can be ignored, is as silly as saying it holds the answer to all our problems.
These things are always hyped up to be panaceas and there follows the disillusionment after the perpetual naysayers trash them as nonsense, until they are quietly accepted and assimilated to achieve a significant but small fraction of what they were originally hyped up to do. Look at DNA: its structure was discovered in the 1950s and yet scientists are still trying to work out how to use that knowledge to change medical science.
Merging different types of data to make up a complete picture is hugely appealing, especially in consideration of how we understand medical and social problems. Indeed if we can map a mass of different factors around what we are investigating, we may be able to pinpoint what are genuine correlations and what are incidental because of other factors.
Make no mistake, like Web 2.0 et al, it will be used for good intentions and bad intentions. The data will sometimes be manipulated, just like any other form of statistics. Remember lies, damn lies and statistics. But we aren't going to ditch statistics as a way of measuring stuff simply because some people use statistics corruptly. Big data is just another tool in the technological arsenal that people can choose to use in whatever way they see a benefit to.
It may even be used to slay a straw man or two...