• Question: Would you say that using algorithms to observe human activity online (e.g. which posts we like/videos we watch) could perhaps poses a danger to privacy?

    Asked by secretsaudience20 on 31 Dec 2019.
    • Photo: Diana Kornbrot

      Diana Kornbrot answered on 31 Dec 2019:


      Yes, Certainly does present danger to privacy
      Its a danger even if you CHOOSE what to like or not, because algorithms are collecting other information, e.g. who else like what you like, and what do they like. You could find yourself guilty by association

    • Photo: Gary Munnelly

      Gary Munnelly answered on 1 Jan 2020:


      Absolutely!

      There is a famous story about Netflix who released a bunch of information to the public about what people were watching on their website. This was all a big exercise to see if tech nerds would look at the data and find new ways to recommend movies to viewers.

      The data was all anonymised, so everyone’s privacy should have been protected. Except, it turns out there are patterns in what people watch, and people will often leave a comment on a review site somewhere after watching a movie. So when two researchers in Texas looked at the viewing habits of anonymous people and cross-referenced that with reviews on IMDB, they were actually able to figure out who more than half of those people were. What was terrifying was they didn’t really need a lot of data in order to do this. Just a couple of movies.

      Preserving privacy can be really, really tricky which is why monitoring people’s behaviour online without their permission is highly unethical. But that doesn’t stop companies from doing it.

    • Photo: James Bentham

      James Bentham answered on 2 Jan 2020: last edited 2 Jan 2020 2:00 am


      Yes, this is already a very important issue. An example beyond activity online is that protestors in Hong Kong are cutting down lamposts, because they may have cameras inside that can be used for facial recognition.

Comments