Big Data / Algorithms / Algorithmic Transparency (4 Mar):

Frank Pasquale – Black Box Society – chapter 1 (pp 1-11)

  • I really enjoyed the metaphor used in the very beginning. That the light is the powerful source that he has to work around. Comparing that kind of power to big data. The fact that nobody knows too much about big data is the reason why it is so powerful. This article argues that this “knowledge problem” is probably for good reason, “to whose benefit”. For pharmaceuticals, they mention that they are allowed to hide the dangers of a drug. Everything we do online is recorded, our credit, our phone location, so where does that information go? For how long? By whom?
  • How can we get that control back? Do we need to continue to rely on whistleblowers? Will companies ever do anything when they know we’re eating at the palm of their hand?

Cathy O’Neill – The era of blind faith in big data must end (Ted Talk, 13m)

  • Algorithm = data + definition of success. She mentions how they take the teachers and scores of their students into an algorithm to shame them, yet nobody outside could access the algorithm. Now how is that fair, when the algorithms could be inherently biased or does not have the full context of information? I also thought the idea of an algorithm of fox news would filter out women because they have not really been the face of success in the past, or most likely people of color as well. This is the type of systemic racism that is put in place today, with police. She suggests algorithm checking through a) data integrity b) definition of success c) accuracy and d) long term effects.
  • How can we allow the public to gain access to these algorithms?

Virginia Eubanks – Automating Inequality (talk, 45m)

  • It’s interesting how throughout the years, even though we change the algorithm, we still end up being racially biased. First it was containment (1819), then investigation (1873), and then digital surveillance -> prediction (1973) which we still use today to keep POC, specifically black people, in their place. Because even with good intentions, they can still have bad outcomes.
  • Have algorithms ever prevented white men from doing anything?

Janet Vertesi – My Experiment Opting Out of Big Data…  (Time, short article)

  • This kind of experiment is really interesting to see just how long you can keep a secret from big data. The internet, other people’s comments, pictures, credit cards, cookies, phone tracking, messages, etc. What I thought was a really powerful comment was the idea that “No one should have to act like a criminal just to have some privacy from marketers and tech giants”. We shouldn’t be required to give our personal information in order to not be perceived as a criminal.
  • Will big data just get even bigger? Is there no way to minimize it?

Walliams and Lucas – The Computer Says No (comedy skit, 2m)

  • The reason that this can even become a comedy skit is the fact that this kind of thing is so relatable and happens so often. We don’t question the system even when the system is clearly wrong. Instead of fact-checking it, we’re more likely to just (as Cathy O’Neill put it) put our blind faith into it. Yes, this is just a sketch, but life imitates art, and there is a sense of accuracy to what is being portrayed.
  • How can we convince the general public to be more cautious in their trust?

Leave a Reply

Your email address will not be published. Required fields are marked *