What are some of the promises and pitfalls associated with data science and human development?
Describe the promise, pitfalls and ways forward Blumenstock uses as the foundation for his thesis.
Big data can be seen as a crucial item in human development for several countries. Data can be used as a cheaper, more timely alternative to censuses and surveys in the collection of data information. This information can be further analysed to guide plans of action for governments when distributing aid. For example, big data in Africa is being used to generate maps in order to see where there is child malnutrition. In areas of epidemic crisis, information from these data sources, such as digital footprints, may even improve public-health interventions. Information from data can be acquired so quickly that it can be used to track effects of a natural disaster almost immediately afterwards, and by the minute. Thus, data has the ability to give big companies and governments a way to acquire relevant information in order to aid in human development.
However, there are several pitfalls of big data that can be seen as a danger. One pitfall Blumenstock expresses is the lack of validation that is associated with information from big data. Mapping from the use of digital data is based largely on generalizations that may not apply over time, or to different areas. This may lead to inaccuracies in the analysation of this information.
Another unanticipated effect is that big data places a great amount of authority to large establishments such as big companies and governments, rather than the vulnerable. For example, most disadvantaged people tend to be underrepresented, and the least represented, in digital data since they are more likely to not have access to smartphones than more wealthy people. In addition, if people become aware of how personal data is monitored and analyzed, some may act towards a way that gives them the most benefit, making aid less available to those who need it. For example, GiveDirectly, a non-profit organization, targeted household with thatched roofs in order to give them direct cash transactions. Soon, others caught on and started to mimic this roof structure to maximize their chances in this system. Blumenstock concludes that digital data may have biased algorithms that marginalized certain groups.
It is also dangerous how big companies and governments can abuse their power further due to the lack of adequate regulation to protect users’ privacy. Only recently did the United States Supreme Court declare that law-enforcement agencies cannot access data without a warrant. In less developed countries, regulation may exist but enforcement does not. Moreover, in some countries, regulation may not exist at all, creating a large issue in data privacy.