AI and Resilient Job Markets: A Balancing Act?
Oaishik Bhattacharya – Student, Kautilya
Why are conversations on Artificial Intelligence important?
The spectre of artificial intelligence and its impact on job markets loom large as the nascent technology evolves into a gargantuan beast that can leave an indelible mark on the course of human history.
Currently, generative AI is churning out content that bears an eerie resemblance to human work, which has triggered experts to voice their criticisms against the technology and its manifold implications. ChatGPT, Bard, and DALL-E, all examples of generative artificial intelligence, are producing texts and images based on prompts or dialogues given by users. Hailed as the state of the art of machine learning, the technology is being extensively used in a diverse range of industries, from media and entertainment to healthcare and scientific research.
So what are AI and Generative AI?
Artificial intelligence allows computers to act and respond in ways similar to those of humans. AI uses algorithms to whisk through bundles of information, identify patterns, predict problems, and solve them. They even learn from their own mistakes.
Meanwhile, generative AI is the “next generation of artificial intelligence” that can create new content from scratch. It is poised to revolutionise how we create and interact with the world.
While research on the technology dates back to the 1960s, in the last decade, generative AI has grown and evolved at breakneck speed. While the inception of WaveNet in 2016 reshaped the GenAI landscape, it was only with the development of GPT4 in 2023 that AI transmogrified into a powerful tool capable of advanced reasoning capabilities and creative text formats.
Recently, the potential of AI to displace workers was at the forefront of Hollywood’s labour fight. Egged on by the disruptive influence of AI in stalling contract negotiations, actors and writers mobilised en masse to reclaim their rights.Research by Anton Korinek and Joseph E. Stiglitz has warned that advancement in the field of AI can cause significant job losses or polarisation, thus widening income and wealth disparities.
What does the data suggest, and what impact will it have on jobs?
A report by Investment banking company Goldman Sachs indicates that Artificial Intelligence could “replace the equivalent of 300 million full-time jobs.” Two-thirds of occupations could bear the brunt of being partially automated by AI.
Another alarmist estimate that outlines how artificial intelligence can actually lead to the “extinction of humanity” was recently published on the webpage of the Centre for AI Safety. It was signed by AI scientists and other notable figures.The advent of AI has triggered anxiety and stress across the world about the possible implications it can have on the job market.
How have the governments responded?
While governments in China and Germany have invested heavily in AI manufacturing, other countries like Italy and Ireland are monitoring the technology and exploring the scope for regulations. Chinese and German companies are developing artificial intelligence robots, which might eventually replace factory workers. In a recent development, a factory in Dongguan City, China, recorded a 250 percent increase in productivity after it replaced 90 percent of its human force with 60 robot arms.
Governments’ eagerness to adopt AI may backfire, as the technology can lead to unemployment and civil unrest, which can directly affect the government’s tax revenue.A McKinsey Global Institute estimate predicts that around 15 percent of global workers, which amounts to about 400 million and 800 million jobs, might be displaced by 2030.In India, 69% of Indian jobs are threatened by automation, suggests a Forrester report. Given the abundance of evidence about the detrimental effects of AI, several countries have now become increasingly fearful of rapid advances in the field.
To assuage fears and calm jittery brains, it is important to note that a seminal study by Somers, Theodorakopoulos, and Hötte highlights how new technologies simultaneously create new jobs if “workers are needed to use these technologies.”Their research demystifies the fear of technology-driven unemployment and furnishes empirical evidence to suggest that the labour-replacing effect of technologies is muted by the “labour-creation” effect.
Another positive estimate by the World Economic Forum suggests that while AI might take away 85 million jobs, 97 million new jobs will be created in fields such as big data and digital marketing. Sobering reports, such as the one published by the MIT Task Force on the Work of the Future, present an optimistic worldview. It highlights how people and computers will work as “super minds” rather than leading to mass unemployment.
What are the jobs that will be affected?
Evidence suggests that tech jobs, such as those of computer programmers, coders, software engineers, and data analysts, will be affected by ChatGPT and other forms of AI. A recent report by Semafor reveals that OpenAI, the creator of ChatGPT, is already training its AI technology to replace some software engineers Similarly, in the media sector, jobs related to advertising, technical writing, and journalism may be susceptible to automation.Other jobs that might face stiff competition from AI are in the legal, research, financial, and teaching sectors.
What are the policy solutions available to reduce job losses?
Governments will have to undertake a detailed assessment of the employees most vulnerable to job displacement by automation. It should identify the sectors that are especially susceptible to job losses and allocate resources to invest in the education and training of such employees. It should also understand the age, gender, geography, and socio-economic conditions of the groups most vulnerable to automation and design policies accordingly. Governments should provide proper incentives and support to workers undertaking transitions to new job opportunities.
Governments can employ tax-benefit schemes to help those who lose their jobs due to automation. Such schemes are essential for reducing income inequality and poverty. Governments across the world should be receptive to this idea.
Tax-benefit systems are similar to safety nets; they provide financial support to vulnerable individuals, or in this particular case, to those who lose their jobs due to automation. The social safety nets provided by the government must be well-funded and capable of providing a justified economic stimulus to the unemployed.
Tax-benefit systems generally consist of two parts: taxes and benefits. Here are a few possible solutions on how such safety nets can be funded:
Taxes are levied on individuals and businesses. In this scenario, taxes can be levied on tech conglomerates. A tax on the purchase and use of automation technologies will help fund and support programmes for displaced workers and foster economic equality. This is basically a redistribution exercise.This approach denotes a direct tax on the purchase of automation technologies. Here, taxes are levied on businesses that purchase robots and automated tools to replace the labour workforce.
Taxes can be levied on the profits of big businesses that use automation on a large scale. Ensuring that the benefits of automation are distributed across strata of society is non-negotiable. Similarly, progressive taxation can be implemented for businesses that use automation indiscriminately.
This taxation will reduce the speed at which automation technology is being introduced across various industries. It will provide governments with the requisite time to build up systems and welfare schemes tailored to deal with the issue of structural unemployment.
As AI continues to evolve, governments and humans should tread cautiously and be prepared to navigate the complex web of challenges and opportunities that it holds for the future.
*The Kautilya School of Public Policy (KSPP) takes no institutional positions. The views and opinions expressed in this article are solely those of the author(s) and do not reflect the views or positions of KSPP.