Politics in AI
Introducing AI: What’s new? What’s different this time?
Artificial Intelligence is the new buzzword. Any moderately aware person of our times has a basic idea about what AI is. Still, as a refresher, AI is an emerging discipline that attempts to create machines who can replicate human-like thinking. Till the advent of AI, human beings did boast of at least one thing that is exclusive to humans i.e. intelligence. This is the basic thing that is being redefined by AI.
To start with, it is important to understand the difference between Automation and AI. Automation is the replication of some action for which rules are fed into the system. The machine doesn’t actually apply its “own mind”. On the other hand, AI is an innovation in which the machine “learns as it works”. The machine collects information, processes it and produces results. Now, this result is fed back into the system and adds to the learning/wisdom of the machine. When new data is added, the machine “reflects back” on the previous results produced. This becomes an iterative process and makes the machine wiser with time. Another thing that is important is that for the first time in human history, disruptive innovation is entering into thinking domain. All other previous revolutions were limited to easing off of manual labor. This is what creates both promises as well as challenges for our times.
Promises:
There are many areas of work where human engagement is either inefficient or is against the moral worth of human beings. In such areas like medicine, retail, banking, marketing etc., AI offers immense possibilities. It can reduce physical labor, improve precision, increase speed, etc. Hence, it paves the way for human beings to engage in more fruitful work. The advocates of AI argue that revolutionizing AI will lead to more leisure and creative involvement of humans. Because of its speed and accuracy in processing information, AI will increase our capacity to process reality and make predictions. These predictions in different domains like politics, economic, public policy, etc. will help us make our lives better.It is redefining human experience by bringing the consumer in the center. The choices of a person get tapped as the data by the program and an algorithm is developed based on it. As more and more people use the program, it develops more knowledge about people. Based on this pool of data, it can suggest personalized recommendations about products.
Concerns:
Despite various promises that AI holds for us, there are many more questions that await precise answers.
First of all, the popular perception about AI is of three kinds. There are people who are advocates of AI, others are opponents while rest and the majority are agnostic. Some vehemently believe that AI will make our lives easier by taking upon tough manual labor as well as unnecessary calculations. Others fear that it will wipe off blue-collar work and create huge unemployment. Others lack the ability to engage with this debate around AI. Hence, there is a “knowledge hierarchy” involved. Not all can participate in this decision making about their upcoming future. There is no restriction formally on anyone but because of knowledge required to engage in this debate, this becomes anti-democratic.
The second important political question involved here is related to geopolitical competition. It was said till now that whoever controls oil can control the global economy. Oil or energy was the main source of power. But with the rise of AI, data has become the new oil. It implies that whoever controls data and puts it to use to develop AI will be the next world leader. This doesn’t mean that this AI revolution gives all the nation-states of the world equal opportunity to compete, innovate and become the world leader. The world presently is also not equal. This gives an advantage to countries like US, Japan, China, etc. This will make the AI-led world a highly unequal place to live.
While talking about political questions around AI, we cannot ignore that traditional notions of the sovereignty of the state in the international arena doesn’t hold that weight. Data doesn’t regard geographical boundaries. Also, there is not an evolved architecture that regulates the flow of data at the national level. This data-driven economy is being dominated by multinational giants like Apple, Facebook, Microsoft, etc. to regulate them is not so easy for weak states especially in a world order guided by the philosophy of market fundamentalism. The threat of monopolization of the economy is imminent.
This dilution of political sovereignty also brings into question the significance of political elites. The rise of AI has given rise to a new techno-elite. We must remember that in a liberal economy, the state cannot interfere much to control the economic behavior of people. But given the power that can be accumulated by a few giant players in the market, it is important to underline that they are not accountable to people. So it needs to be seen how elected representatives hold techno-elites accountable.
Another important axis to look at AI is through its interaction with our existing criminal justice system. In case of accident and loss caused by some program, it is not clear who will be punished. For instance, if a driver-less car kills a person, who will be accounted for? If some person suffers loss because of some program, who will pay for monetary compensation? Will it be the programmer or the owner of the company? There are live questions that need to be addressed.AI will also pose challenges to existing economic model i.e. Capitalism. If we agree that there will be job losses due to automation and the rise of the robot economy, the question rises where demand will be generated for the consumption of products. A suggestion is being floated about giving all a universal basic income but its design is not yet clear. It is also noteworthy that AI will concentrate wealth in the hands of a few people and render the mass redundant or irrelevant. This may create a cultural upheaval in the society where alienation may make people suffer. There can be a sense of uprootedness. In essence, the very idea of ‘’human condition’’ is going to be redefined.
Finally, the question remains if robots will be able to do value judgments. Other than intelligence, which also sets apart human beings is the capacity of moral judgment. Since data cannot capture values and idea of good, it is not clear how robots will engage in ethical reasoning. In its absence, we are doomed to be dominated by instrumental reasoning.
Conclusion:
We see that the production of AI is imbedded with political questions in the sense politics decides who gets what, when and how. It suffers from old societal concerns like inequality, hierarchy, and accountability. It also divides as it claims to unite. It has the potential to divide society into super-rich and redundant human beings. Our journey to a robot led world is fraught with scapegoatism of many people whom we will be asking to sacrifice and take the pain of job loss as future generations allegedly benefits. Those who will be deciding our fate need to be held accountable by whom we elect. We need redesigning of institutions for new challenges. The costs and benefits of producing and consuming AI need to be first calculated clearly and then distributed justly.