Michael Battalio


Sunday, April 05, 2015

Serious conversations (part 60): Data collection and surveillance

        This series is a continuation of my conversations with an atheist friend of mine. These are my edited responses from that conversation. The sixtieth – sixty-second entries deal with data surveillance.

        From my discussion on government surveillance (see My politics parts 4 and 5) it was noted that there can be benefits to data collection, assuming that the collector can be trusted. An example could be store cards offered by grocers and pharmacies that give discounts for logging all items purchased. From that idea I draw several difficulties.
       
        There certainly are benefits to data mining. With large amounts of user data, companies that develop public products can quickly improve them. The anecdote that comes to my mind is with software updates. I've repeatedly heard the adage that you should never be part of the first group of people to adopt a technology because no matter how much lab testing a piece of software or hardware has gone though, the engineers simply are not going to be able to duplicate all possible situations that a customer base of millions will put a product through. You wait to adopt until after the first release that way all the early adopters will have sent all of their bug reports in, and the engineers will have fixed the unanticipated problems.
        I think that many companies genuinely have their customer's interests in mind if for no other reason that if the customers disapprove of tactics employed by a company (assuming knowledge of shady or underhanded tactics is revealed), they will switch to a competing company. Additionally, a trained analyst may be able to anticipate what we want before we know or be able to discern our desires better than we can. However, we need to specify the difference between collection of data with our express consent and collection without our consent or worse collection after we explicitly deny permission. Each of those possibilities occur.

        There are two problems with data collection via a store card. Both are rather paranoid ways of looking at the world but worth mentioning. The first problem is that even though there are a lot of people in the word, so it is very difficult to track people, each of us is rather unique if enough datapoints are collected. What I mean is if some nefarious entity were able to access your Kroger card information with all of your buying habits, they could at some point in the future dramatically narrow down where you are later assuming consumption habits would later be determined (not necessarily though another store card). Only so many people are going to buy the specific combination of products you frequently buy. This situation is highly unlikely, and the only circumstance I can conjure up where such paranoia could be useful would be if you had to go into a witness protection program. What I'm saying is that every piece of data you surrender (either willingly or unwillingly) becomes a method of tracking you later. Us being regular boring citizens, this is never going to be a problem especially since grocery/pharmacy store cards are opt in.

We cover the second problem in the next entry.

No comments:

 
2003-2016 Michael Battalio (michael[at]battalio.com)