AI with its exceedingly powerful algorithms is already here and plays an increasing role in shaping our societies, economies, and politics. The content we absorb on Facebook, movies we watch at Netflix, the music we listen to on Spotify rely on predictive modelling by algorithms. These algorithms learn to save the feed on the basis of personal choices for marketing our experiences. They are not only used by social media platforms but also used by governments all over the world in the public sector i.e., education, healthcare, tax administration and criminal justice. For example, algorithms can be found being used by judges who decide the reoffending of a criminal or not.
We need to understand that either teachers use algorithms to assign assignments to students or algorithms are used by municipalities to figure out prospective routes for waste collection; we are living in an algorithmic society. Most importantly, the influence of algorithms in day-to-day lives refers to the application of advanced statistical methods and techniques used by private entities to automate decision making on people’s choices either which brand to choose or which party to vote for. Thus, transforming the society into an algorithmic society in which large social media platforms work between nation states and individuals and use of AI and algorithms to govern populations.
There is mind-blowing amounts of data which is being generated about our daily lives with algorithms processing and making decisions which control and nudge our behaviors. This not only further expands the possibility of control and surveillance by nation states but also opens up a new paradigm associated with the closure of decision making by individuals and shift in the functioning of power. Decision making by humans are being replaced by AI, databases, and automated procedures. The discretionary shift to IT professionals who design algorithms, data analysts who identify behavioral patterns and algorithms themselves which recognize new patterns and settle their decision-making process through machine learning is immense. Over the past few years, the power of algorithms was felt widely and it should be treated as a form through which power is enacted in our society. Undoubtedly, the power of algorithms lies not only in computer codes but also in the way they are developed and designed. It is inevitable to analyze how algorithms are reshaping the relation between nation-state, citizens and technology companies.
Scholars have tried to conceptualize and define the power of data in an extension of Deleuze’s writings. They have defined data in terms of ‘info power’ (Koopman), ‘data power’ (Chamoyou), ‘expository power’ (Harcourt) or ‘psycho-power’ (Stiegler). It means that not only the impact and repercussions of algorithms need to be identified but also the powerful way in which they oscillate in the social order processing. There is no second thought that algorithms are more efficient than humans, but we need to see whether they are fairer or not. Proponents claim that by obliterating the wicked nature of humans and its impact on millions of people, algorithms can make decision making less prone to biases and sentiments. However, critics are of the opinion that algorithms usually behave in a way which reflect various patterns and prejudices deeply immersed in society. Though algorithms lack conscious racist intention, but they can regenerate or propagate certain values which reflect the biases of system designers. For example, there is increasing concern that ‘dirty data’ used in policy practices, taken from racially based, flawed practices eventually lead to discriminatory results. This also holds correct for algorithmic decision making which is based on machine learning. Another important corroboration against the efficacy of algorithms is associated with privacy and data protection. All those systems developed using AI pose significant privacy risk. Governments all over the world need to think deeper about how trustworthy AI systems can be and how fairer algorithms behave by looking at both risks and solutions associated with them.