Ted Talks: The Era Of Blind Faith In Big Data Must End
Data skeptic, Cathy O’Neil, while giving her inspirational talk uncovered pretty much regarding the dark secrets of the big data. She outlined how the humanistic algorithms could reinforce human bias. Those algorithms usually determine who gets a loan, who gets insurance, who gets a job interview and much more.
But the truth of the matter is that that they don’t make things fair automatically. The speaker believes that they are often far from the various scientific concepts. O’Neil came up with a term for algorithms that are important, secret and harmful: “weapons of math destruction.”She calls upon everyone to try and learn something about the various hidden agendas behind the supposedly objective formulas and why people should consider establishing better ones.
To build an algorithm, one needs data and a proper definition of success. The data refers to those things that happened in the past whereas a proper definition of success has to do with what one has actually been hoping for. A lot of people don’t formalize algorithms in written code. However, that doesn’t mean they don’t use them. In fact, everyone uses algorithms.
The speaker admitted that each day she used an algorithm. She gave an example of the meals she prepared for her family. She referred to the ingredients in her kitchen, her ambition and the time she directed in the cooking as the associated data. However, she didn’t consider the little packages of ramen noodles to be food.
The first rule of algorithms refers to the aspect of one being in charge. Someone needs to get the feeling that his opinion matters.
People think of algorithms from different dimensions. Some observe them as objective, true and above all scientific. Analysts see that as a marketing trick that aims to intimidate people so that they end up trusting and fearing them. It has something to do with taking advantage of the fact that a lot of people fear mathematics. Whenever humans choose to direct a lot of blind faith in big data, quite much ends up going wrong. Algorithms can be described as the various opinions that are embedded in a code.