Labour and Artificial Intelligence: Visions of despair, hope, and liberation
In the United States, job demographic data from censuses since the 1900s reveal a startling fact. Despite the two post-Industrial revolutions of electricity and computers, the occupations with the largest employment numbers are still jobs for drivers, retail, cashiers, secretaries, janitors etc, i.e. old professions needing simple skills and mostly repetitive work. This lack of transition to “newer” jobs is a global phenomenon, especially in the global south. India, for example, has half of the working population doing agriculture.
One must grasp the significance of Artificial Intelligence (AI) in this context. Unlike technological upheavals of the past, AI is unique in that it can rather cheaply replace a vast spectrum of mental, creative, and intuitive human labour. With AI presiding over the mass extinction of repetitive jobs, precisely the sort employing the most workers, no precedent exists of newer jobs replacing them in large enough numbers.
There is no dearth of alarmist narratives around AI. But the “danger” of AI isn’t that it will become hostile, or follow its instructions with such a literal interpretation and on such a scale that human existence itself be jeopardised. Science-fiction scenarios of rampant AI are interesting thought-experiments but already-existing AI is here, and requires well-crafted policy.
Take driverless cars. The US Bureau of Labor Statistics lists six million professional drivers in US as of 2016 and their jobs are in peril. Trains are easier to automate, and metros like Rome, London, and Paris to name a few are already transitioning. An MIT Tech Review 2016 article describes factories, warehouses, and farms developed in China needing minimal humans to operate. A US firm, WinterGreen Research Inc, projects that agricultural robots can become a $16 billion industry.
In India, Maruti Suzuki India Ltd already has one robot for every four workers at its Manesar and Gurgaon factories. McKinsey reports in 2017 that half of the Indian Information Technology workforce will become “irrelevant” in four years. Some industry watchers advocate retraining in social skills, under the prevalent but incorrect notion that machines cannot replicate human empathy and genius. AI, however, can perform “creative” or “empathic” types of labour. Caregiver robots will eventually enter nursing. The arts (including music) are subjects of AI research with Artificial Creativity as a subfield. Even areas like journalism, teaching, and entertainment are not entirely immune. Sophisticated processes like answering free-form questions in natural language is being actively researched, and will dramatically change the service sector.
In medicine, auxiliary work is easy to automate but the real challenge arrives when AI starts making better diagnoses than humans, which was demonstrated for cardiovascular diseases in an April 2017 paper. In the field of law, interns and junior lawyers, the backbone of legal firms, doing tasks like “discovery”, can be replaced. Finally, US banking giants like BNY Mellon, BBVA, and American Express, have spent hundreds of millions of dollars on AI research, and low-end banking jobs might get axed.
A 2016 study by Deloitte states that 35% of jobs in Britain are at high risk in the next two decades. A 2016 McKinsey report pegs the potential for automation in US at 75% for food services, 40% in services, 35% in education, and 30% in administration. And unlike the first world India lacks the robust welfare state to support our underpaid contractual labour when automation hits our shores.
Given that Artificial Intelligence is revolutionary, and imminent, ‘what is to be done?’
The worst policy is to do nothing. A broken labour market alongside the euphemistically named “sharing” economy wherein monopolies own vast assets managed by AI and people only rent (think Uber like services not just for cars but for everything, operated via AI), presents a real danger of a regression to a system where only capital is needed and most of labour isn’t. Futurists call this “neo-feudalism”. In an extreme case, much of the working population might become irrelevant to the economy, reduced to penury, and locked out of civilised sustenance.
A panacea which technocratic thought leaders are advocating is Universal Basic Income (UBI). The idea behind UBI is that we accept unemployability of most humans and to preserve a minimum living standard and consumption needed for the economy, all welfare measures be streamlined, and a regular quantum of money paid to everyone. This money could be, as Bill Gates suggested, obtained via the heavy taxation of automation, or might come from public wealth, like land or oil.
The problem is, in a trivial form, UBI does nothing to address the root cause — the control of vast productive forces by an ever-decreasing few. It gives up on most of the population, relegating them to an infantile, consumerist role, to be only fed and entertained, with no chance at social mobility. It does noting to correct the pseudo-scarcity the market creates in an otherwise era of AI-led hyperproduction. This is a waste of the potential of both humanity and AI.
There is another way, however, which doesn’t treat AI as a technological artefact separate from socio-political forces, but as a component of public policy. AI could be a public good, and not merely an awe-inspiring private resource that companies can dazzle us with. There’s a need to challenge “free-market” fundamentalism and initiate international cooperation on AI policy, and start large, public funded, and distributed AI research, AI public-works, and AI-centric education.
In the light of this “who controls the AI?” becomes a vital question. A serious conversation is needed, especially in the third world, on how the AI led production of the future be managed: Will it be democratically controlled, or driven by corporate shareholders? AI can conceivably and radically improve both distribution and productivity, augmenting individual and public affluence. In other words, AI need not be market driven; there is a case for conceptualising it as a public good that is used to realise a better redistribution and a critical tool for shaping and strengthening democratic institutions.
AI can upend the metaphorical gameboard and liberate labour. It is a historical opportunity.
Anupam Guha is with the University of Maryland
The views expressed are personal