Hello, I wanted to do something different today. Recently i applied to a data science masters at a local university and for the application, you needed to discuss the ethical and societal implications of an aspect of Data Science. Unfortunately, I didn’t get accepted on the course so I’m going to share it with you as I thought it was an interesting question.
“Information is the oil of the 21st century, and analytics is the combustion engine”. However, with this great power comes with great responsibility.
Data analytics encompasses everything from daily shopping habits, to application usage on a smartphone and, more significantly, to users’ Google search history. The data collected, generated and analysed can have huge political ramifications in determining election outcomes and, ultimately, party victories or losses.
Nonetheless, this begs the question of how ethical is such data harvesting. In particular, there are key ethical debates surrounding awareness and understanding of each and every individual who is targeted in the name of Data Science. An especially challenging issue of such collection analytics is that very often data is generated without an individual’s consent, knowledge or understanding. Practices- that if occurred in other fields- would be viewed as highly unethical.
Algorithms are now more and more prevalent throughout society. These can be extremely useful in some situations, from suggesting suitable music linked to a user’s profile, to presenting advertisements in line with a Google Search history. Nevertheless, if the algorithm goes wrong in these instances, you either end up listening to a song for 30 seconds before moving to the next one or see an advert for something you have no interest in. Furthermore, other algorithms, such as ones being used to decide on whether an individual gets parole, can have a far more sinister outcome. In order to be considered ethical and fair, these life-altering algorithms must, in my view, have some form of independent verification. This is to ensure they are just and free from discrimination to all ethnic groups, sexual orientations and political affirmations within society. The users of these algorithms should be thoroughly trained with the understanding that no decision the algorithm can produce can be 100%. In other words, there will always be an element of probability.
Significantly, the Cambridge Analytical scandal is one of the key ethical case studies within data science. The negative media coverage and public outcry from this scandal suggests that it is highly unethical to obtain an individual’s data from a public website and, subsequently, employ this sensitive information to target and direct political advertising and influence party voting. However, would such data harvesting still be perceived as unethical if it was employed in medical research? For example, what if the data had been utilised to identify people in the earliest stages of cancer and could, most positive of all, increase survival rates? Would there be such universal outrage and, perhaps, a more ethical stance taken if the data was to be employed for “good”?
Additionally, data science is also having an impact on society with the risk of improperly communicated data being used to reinforce a particular belief and, ultimately, as a weapon to heighten prejudice. This has a wide impact, particularly in general elections and could cause politicians to focus on the extremely misguided policies. This could lead to the requirement for an independent body within a society which will fact check and review all published data science work to ensure the public receive accurate information.