The historical evolution of Artificial Intelligence (AI) dates back to the year 1996 when Deep Blue AI defeated the then world chess champion. Garry Kasparov The year 2019 witnessed geopolitical paradigm where there was a race for technological supremacy between superpowers. It is estimated that by 2034-40, 50 per cent of the jobs would be automated in United States i.e.; within the next 15 years (Lee Kai Fu, AI Superpowers).
Also majority of researchers predict singularity by 2045 — a stage where machines become more advanced than human beings. This necessitates one to understand AI, its benefits, its major issues and its implications on government and social order.
Scientists have defined AI technically and its relevance to the global technological revolution. In simple language, it is the technology that can think and act like humans, think rationally, or act rationally or the technology that simulates or imitates human intelligence, but with super-charged processing power.
AI high end applications may be broadly categorised into three parts — firstly, machine learning; secondly, natural language processing; and thirdly, predictive analytics.
Machine learning finds key relevance in fraud detection and surveillance which may include pedestrian detection and traffic light control, automatic license plate reading for access control, for public safety. The facial recognition technology coupled with crime-busting is another revolutionary application which provides for twenty-four hours of image processing across various sectors.
As far as ‘natural language processing’ is concerned, its market is set to grow to $22.3 billion by 2025 with wide ranging application across customer service, healthcare documentation and voice-control environments. The predictive analysis of AI finds its relevance in areas such as hyper-personalised marketing, social impact marketing, wherein, the same may be helpful for diabetes prevention, educating anti-vaccine communities, targeting child pornographers, safe sex and condom usage etc.
However, the key question is the challenges of AI to policy makers which may result in large scale disruption causing shifts in society. The three board issues for consideration in this regard may be (a) the future of work, (b) data and privacy, and (c) ethical governance.
The concept of workplace may also be redefined with the advent of AI with the definition of workplace incorporating global platform with 24/7 concept, work at home as also freelancers, resulting in reduction of office buildings, increase in small office home office (SOHO) which is similar to the practices in retail industry. AI may also have ‘labour’ implications resulting in job displacement wherein studies reveals that application of one robot may displace 6.2 workers.
Fifty one per cent of the jobs in OECD countries may be automated as estimated by the World Bank and 47 per cent of employment in the US may be at risk as predicted by University of Oxford. The net labour force implications may be job growth for better educated and skilled personnel coupled with lowest employment rates for the blue-collar workers.
The social implications of AI include redefining the concept of social safety nets, social investment stipends based on social productivity as suggested by AI expert Lee Kai-Fu, due to proposed challenges on social values like love and compassion with the advent of AI.
The financial implications on the other hand may be the emergence of new frameworks of taxation skewed towards company tax with less individual taxes. The interesting question that needs to be studied is of how to tax robots and freelancers. The legal implications may be that if AI is wrong, who is to be blamed; whether the human expert or the company who bought it or the developer or the programmer? The code of ethics for professional robotics engineers subject to research and application in European Union may be relevant in this regard.
The implications of AI on education sector may lead to its revamp with programmes such as ‘Skills future’ developed in Singapore in this context, wherein its concept vis-a-vis reality may be interesting. The implications on employment laws may involve exploring concepts such as quota for human workers, working hours for freelancers as boundaries of private and work life being blurred, requirement of new labour and employment legislation with increased automation, etc.
The Universal Basic Income concept as floated by Lee Kai-Fu may be interesting in this context. The key dilemma in this context for the developing countries may be limited financial resources, more unskilled workers coupled with higher unemployment for blue collar workers, which may aggravate poverty issues.
The protection of personal data in the AI environment may be a serious challenge wherein there may be trade-off between privacy and prosperity. The General Data Protection Regulation in European Union, the sectoral and state laws in US and cybersecurity law in China may be taken as the basis for policy formulation. The ethical considerations and governance issues of AI may be redefining regulations and governance with focus on fairness, safety, reliability, privacy, inclusiveness, transparency, and accountability.
The challenges in this regard are the inadequacy of the preliminary guidelines and the broader questions which may arise in this regard.
The future considerations of governments with the advent of AI may be its challenges to political legitimacy, redefinition of the social contract and its implementation in the government. The challenges to political legitimacy may revolve around government versus private companies as central provider of solutions, increasing dissonance in citizens on issues like privacy, data and digital rights, requiring legitimacy of the technology in the implementation of AI in government functions, etc.
The redefinition of the social contract may involve the role of government and its regulations, economic well being to social well being, e-legitimacy of AI with consensus and socially accepted purpose, Constitution based on new shared principles, etc.
However, the major challenge for the society may be the changing values, with requirement of more “human” values such as love and compassion, co-existing with AI.
The writer is Deputy Secretary, Ministry of Finance. Views expressed are personal