According to Cukier “we’re just at the outset of the big data era,” so we have to use our imaginations to consider what all of the variable uses for big data in the future will be. But Cukier also points to a powerful, concrete example that we can see today.
Twitter can tell us, for instance, that subpopulations exist that are either immunized for the flu or are not. This insight completely upends the idea of ‘herd immunity’ because “there’s whole subgroups of the population that all don’t get vaccinated, yet they all hang out together,” Cukier points out.
It took datafying relationships and interactions to learn this “deadly important” insight about flu vaccinations, Cukier says. “So when you think about it in the grand scheme, what big data means is we are able to learn things about ourselves at the population level, at a huge scale, that we never could in the past.”
It looks to me too far-fetched to extrapolate data like flu vaccinations performed on populations from social media like face book and twitter.Unless by the use of these media such data are consciously called for by an interested aggregator like the government or non-profits organisation involved in that type of work. Or unless data like phone number, address etc. that may be incidentally available from the social media are used for contacting individuals and collating their particulars.