This article on Human Development with Data Science was very specific and very interesting. I liked the structure of the article a lot, as it wasnt empty promises and wishes. Blumenstock gave what he believes are specific solutions and responses. The Promise is one of the paragraph in which Blumenstock begins his explanation of the different ways to use Data Science to help people who really need it, rather than the majority of humanitarian aid resources going to those who are well enough to survive with out them. Blumenstock uses a few specific examples to prove this. First, he explains how “high-resolution maps” can be created to show which areas have the most need, in regards to child malnourishment and low food production. By using this we could narrow in on which areas need more intense care. This also relates to his bigger goal, to not lose sight of the humans behind big data. This method ensures every human is accounted for and that they aren’t just a number on a screen, they are a picture, and they are a person. Secondly, he explains how in Africa, there are several countries that look at the correlation between the number of international calls made per household, and the roofs of which these people live under, and relate them the the individual weath of that household. In his next paragraph, Pitalls, he descibes the potentially detrimental outcomes. Personally, the two most important being the Biased Algorithms and Lack of Regulation. Blumenstock states that “those who are poorly represented are often marginalized”. If a single algorithm in this process is biased, the whole system contradicts itself. What once was an idea to consider every person needing aid, is now just a number on a paper, or even leaving people out of the program entirely. This leads to the Lack of Regulation. This is a tremendous amount of power, and one which, left in the wrong hands, could go wrong very quickly. In Ways Forward, he gives solutions to the problems he suggests in Pitfalls. He speaks of how it can be regulated properly, and how it can lead to a more collaborated world view. This common goal could be very impactful.
In order to create an impact of those around us who need it, we must make a concious effort to put the person before the number. It is easy to get lost in the statistics, but at the end of the day, these statistics could determine where someone is sleeping or eating tonight. By using this to propel our inovation and growth in Data Science, we can create real solutions that recognize each person and their struggles, and use the numbers given to us by them to create a better, more efficient and resourceful world. To help create an unbiased solution, being “transparent” as Nira Nair put it, will help narrow in our findings significantly. If any ounce of bias creeps into an implemented program, the whole program could be shot. To get accurate results, that can actually help the people intended, transparency is necassary to get the whole picture and to see below just the statistic.
It becomes a very difficult balancing act. Trying to find a solution that not only encapsulates the humans behind the data, but is also able to come up with real-life solutions is not easy. Blumenstock does a very good job of including attainable solutions that recognize the people. Another example of big data and human development intersection he included in his article, was how public health interventions could be improved by tracking different kinds of cell phone data and social media, to find the specific areas that need more intervention after some sort of disaster. This kind of solution takes a special kind of person and program to implement. It would have to ensure no bias, so if family and friends were involved, that would have to be ignored if it weren’t the area that needed the most help. These types of approaches could heavily impact different areas of the world today that are having a natural disaster, such as the fires in Australia.