VANI BANSAL, ANANYA PADMAVATI NIGAM, SIMRAN RAKESH SAMPAT, AND ANUPAMA NAGABHUSHAN RAO – 2024
With their promise of unprecedented levels of efficiency and exponential growth, artificial intelligence (AI) and machine learning (ML) are ground-breaking technologies that have impacted every industry. The term artificial intelligence (AI) describes how computers simulate human intelligence processes, and machine learning (ML) is a subset of AI that enables systems to learn from data without explicit programming.1
AI and ML have a significant and wide-ranging effect on the labor market. Studies have shown that automation and artificial intelligence may displace up to 800 million jobs worldwide by 2030.2 Further, an IBM report states that 59% of Indian enterprises with 1,000 or more employees reported using AI actively in their operations. Early adopters are setting the standard, according to the “IBM Global AI Adoption Index 2023,” with 74% of Indian enterprises already using AI and having increased their investments in the technology over the previous 24 months in areas like workforce reskilling and research and development.3
According to economic theory, AI impacts the labor market in two ways. It can negatively affect the availability of jobs as machines and computers are substituted for labor (displacement effect). However, it may also have positive effects through the increase in productivity (productivity effect) and the creation of new jobs (reinstatement effect).4 These occur simultaneously and their magnitude will determine the overall impact on the labor market.
Historical Context
When talking about the implications of AI in the labor market, it’s important to look at the historical context. ‘Automation anxiety’, the anxiety of massive job losses due to the advent of new technology, has existed for centuries. This was particularly pervasive during the Industrial Revolution in England, when it was believed the advent of new technology such as the spinning jenny and steam engine would lead to large amounts of unemployment. There was some merit to these grievances. Entire industries were now seen as pointless pursuit, and a small increase in real wages—4% from 1760 to 1820—was not enough to increase living standards as infant mortality rose and life expectancy fell in the UK during that time.5 Nevertheless, unemployment remained reasonably constant. While the industrial revolution led to the disappearance of some industries, in the long run, it created far more jobs, increased productivity and lowered prices for goods. This could be due to the increased importance of the positive forces over the displacement effect.
Positive Implications
Through the productivity effect, AI has the potential to drive large increases in productivity by automating routine tasks. According to a study by Wang, Khosla et al, when doctors used deep learning systems to identify metastatic breast cancer, their error rate reduced by 85%, showing the potential productivity gains of ML.6 In other industries, Generative AI can improve a highly skilled worker’s performance by up to 40%.7
AI presents a unique opportunity for economies to leverage capital deepening, particularly developed economies where the integration and accessibility to AI are more streamlined. According to Goldman Sachs, innovations in generative AI can increase productivity growth by 1.5% over the course of ten years and boost the global GDP by 7%, or nearly $7 trillion.8 Workers also stand to gain from potential upskilling, as they train in the usage of AI and improve the scope and value of their work.
As seen in the past, although technology replaced some occupations in the white-collar sector, there was a requirement for new tasks to be performed by labor.9 Labour will also always be necessary for the creation and upkeep of AI and ML systems, creating demand to compensate for the lost jobs. The Future of Jobs Report by the World Economic Forum predicts a rise in demand for AI and machine learning specialists, data analysts and scientists, and information security analysts, leading to the creation of 2.6 million new jobs. Similar trends have been seen in previous revolutions which have led to a change in the composition of GDP to accommodate new technology. Such shifts would have been incomprehensible during these times, and it’s highly likely that as the world adopts AI we will move to a new form of economy with jobs that have not necessarily been seen before.
Furthermore, AI promises to increase employment through improved labor market matching—the process by which workers are matched to jobs. AI presents opportunities for efficiency gains, increasing the equilibrium employment.10
Negative Implications
Some argue that AI is fundamentally different from previous machinery in that it has the ability to automate not only manual tasks but also cognitive tasks once seen as uniquely human. A report by McKinsey Global Institute, ‘Generative AI and the future of work in America’ states that by 2030, activities accounting for up to 30% of hours currently worked across the US economy could be automated.11 This displacement effect can cause a reduction in demand for labor as well as a decline in wages and employment. Further, the efficiency gains from automation leading to increased output per worker do not necessarily translate into a corresponding increase in demand for labor. This causes a decoupling of wages from output per worker and consequently a reduction in the share of labor in the national income.
AI will also have a significant impact on income and wealth inequality. While the risk of displacement due to AI extends to higher-wage earners as well, we must also consider that potential AI complementarity has a positive correlation with income. Hence, the effect of AI on income inequality will depend on whether AI complements high income workers or displaces them. If they are complemented, this would lead to a large increase in their income, and hence an increase in inequality.
Image Source: 2024 IMF report, ‘Gen-AI: Artificial Intelligence and the Future of Work’
As seen in the graph, approximately 60% of jobs in advanced economies, 40% of jobs in emerging market economies, and 26% of jobs in low-income countries are susceptible to AI.12 Thus, while developing economies might experience fewer immediate disruptions from AI, they are also less prepared to harness its benefits, deepening the income disparity.
Therefore, while the argument exists that the AI revolution will follow in the paths of all other previously ‘groundbreaking technology’ it’s also important to be cognisant of the differences between the AI revolution and previous industrial revolutions while discussing its implications on the labor market.
Potential Economic Policies
In order to discuss policies geared towards AI, it is important to take into account the AI Preparedness Index by the IMF which considers parameters such as digital infrastructure, innovation and economic integration, human capital and labor market policies, and regulation and ethics.13
A policy of education and training, integrating AI into the formal education system at all levels would be the most universally applicable approach. While many universities have already started offering AI-related courses, it will be beneficial for school children to learn how to efficiently use AI and ML from a younger age. Advancements in AI transform the skills required in the workforce from routine cognitive tasks to more interpersonal and creative abilities. Education policy should pivot to include more analytical and creative thinking, interpersonal communication, and emotional control.14 Governments should also invest in jobs that prioritize these skills.
Further, it is important to inculcate an environment of continual learning and upskilling into adulthood to keep up with rapid technological changes. To incentivise this, governments can provide credit or tax-free funds to citizens that are upskilling, as Singapore has done through the SkillsFuture programme.15 They should also provide a social safety net to those experiencing technological unemployment.
One concern with AI and ML programmes is their use of large amounts of data, often including copyrighted material, without the owners’ permission. This creates the need for strong intellectual property laws aimed at protecting the creative property of artists and minimizing job losses within the creative industry. Another ethical consideration is the privacy of both consumers and employees interacting with AI. Strong laws must exist to protect personal information of all stakeholders as the information collected is non-rivalrous and cheap to store, while having negative externalities.16 It is also necessary to discuss the question of liability. We propose that the manufacturers of the AI should be held liable. While this may slow down innovation, it will incentivise responsible use of AI and minimize distortions in the labor market.
According to the AI Preparedness Index, countries that lack foundational AI preparedness, generally low-income developing economies, should focus on digital infrastructure and human capital to enable the smooth adoption of AI. On the other hand, countries with a strong foundation, such as high-income developed economies, should focus on supporting digital innovation and creating legal and ethical frameworks around AI. International agreements like the Bletchley Declaration17 and UNESCO Ethics of AI18 will ensure uniform adoption of these frameworks.
Additionally, policies can be developed to reduce the relative cost of labor as when labor is comparatively cheaper, firms are disincentivised to replace labor with AI. For instance, strategies like capital taxation, which levy a tax on AI-related capital, can make labor more cost-effective.However, this could lead to the issue of capital flight. To resolve this, international agreements may be used to ensure a uniform tax, preventing countries from bearing an undue burden.
In conclusion, it is clearly evident that AI has the potential to significantly disrupt the labor market. While it has garnered a lot of attention for its potential negative impact on job availability, one must not ignore the possibility of large gains in productivity. Policies should be geared towards amplifying these positive forces and dampening the negative forces, while
ensuring equitable access and implementation of AI globally.
Bibliography:
1. CU-CAI. “Artificial Intelligence (AI) vs. Machine Learning | Columbia AI,” n.d.
https://ai.engineering.columbia.edu/ai-vs-machine-learning/.
2. Manyika, James, Susan Lund, Michael Chui, Jacques Bughin, Lola Woetzel, Parul Batra, Ryan
Ko, and Saurabh Sanghvi. “Jobs Lost, Jobs Gained: What the Future of Work Will Mean for Jobs,
Skills, and Wages.” McKinsey & Company, November 28, 2017.
https://www.mckinsey.com/featured-insights/future-of-work/jobs-lost-jobs-gained-what-the-future-of-work-will-mean-for-jobs-skills-and-wages.
3. IBM India News Room. “59% of Indian enterprises have actively deployed AI, highest among
countries surveyed: IBM report,” n.d.
https://in.newsroom.ibm.com/2024-02-15-59-of-Indian-Enterprises-have-actively-deployed-AI,-highest-among-countries-surveyed-IBM-report.
4. OECD Employment Outlook 2023. OECD Employment Outlook, 2023.
https://doi.org/10.1787/08785bba-en.
5. Wang, Li. “Daniel Susskind (2020). A World Without Work. Dublin: Penguin Books.” Nuevas
Tendencias, no. 108 (June 24, 2022): 47.
6. https://doi.org/10.15581/022.43221.
7. “How generative AI can boost highly skilled workers’ productivity,” n.d.
https://how-generative-ai-can-boost-highly-skilled-workers-productivity.
8. Goldman Sachs. “Generative AI Could Raise Global GDP by 7%,” April 5, 2023.
https://www.goldmansachs.com/intelligence/pages/generative-ai-could-raise-global-gdp-by-7-percent.html.
9. Goldin, Claudia, and Lawrence F. Katz. “The Race Between Education and Technology: The
Evolution of U.S. Educational Wage Differentials, 1890 to 2005,” March 1, 2007.
https://doi.org/10.3386/w12984.
10. Blanchard, O. et al. (1989), “The Beveridge Curve”, Brookings Papers on Economic Activity,
Vol. 1989/1, p. 1,
The Beveridge Curve.
11. Blanchard, Olivier Jean, Peter Diamond, Robert E. Hall, and Janet Yellen. 1989. “The Beveridge
Curve.” Brookings Papers on Economic Activity 1989 (1): 1. https://doi.org/10.2307/2534495
12. Ellingrud, Kweilin, Saurabh Sanghvi, Gurneet Singh Dandona, Anu Madgavkar, Michael Chui,
Olivia White, and Paige Hasebe. “Generative AI and the Future of Work in America.” McKinsey
& Company, July 26, 2023.
https://www.mckinsey.com/mgi/our-research/generative-ai-and-the-future-of-work-in-america#/.
13. Tavares, Mauro CazzanigaFlorence Jaumotte, Longji Li,Giovanni Melina,Augustus J
Panton,Carlo Pizzinelli,Emma J Rockall,Marina Mendes. “Gen-AI: Artificial Intelligence and the
Future of Work.” IMF, January 14, 2024.
https://www.imf.org/en/Publications/Staff-Discussion-Notes/Issues/2024/01/14/Gen-AI-Artificial-Intelligence-and-the-Future-of-Work-542379?cid=bl-com-SDNEA2024001.
14. Agrawal, Ajay, Joshua S. Gans, and Avi Goldfarb. “Economic Policy for Artificial Intelligence.”
Innovation Policy and the Economy 19 (January 1, 2019): 139–59.
https://doi.org/10.1086/699935.
15. SkillsFuture SG. “SkillsFuture Credit,” n.d.
https://www.skillsfuture.gov.sg/initiatives/mid-career/credit.
16. Street, Prime Minister’s Office 10 Downing. “The Bletchley Declaration by Countries Attending
the AI Safety Summit, 1-2 November 2023.” GOV.UK, November 1, 2023.
https://www.gov.uk/government/publications/ai-safety-summit-2023-the-bletchley-declaration/the-bletchley-declaration-by-countries-attending-the-ai-safety-summit-1-2-november-2023.
17. “Recommendation on the Ethics of Artificial Intelligence,” n.d.
https://www.unesco.org/en/articles/recommendation-ethics-artificial-intelligence.
18. Acemoglu, K. Daron, and Pascual Restrepo. “Automation and New Tasks: How Technology
Displaces and Reinstates Labor.” the Journal of Economic Perspectives/the Journal of Economic
Perspectives 33, no. 2 (May 1, 2019): 3–30.
https://doi.org/10.1257/jep.33.2.3
19. “Toil and Technology — Finance & Development, March 2015,” February 26, 2015.
https://www.imf.org/external/pubs/ft/fandd/2015/03/bessen.htm.
20. Acemoglu, Daron, Pascual Restrepo, David Autor, Toulouse Network on Information
Technology, Google, Microsoft, IBM, and Sloan Foundation. “ARTIFICIAL INTELLIGENCE,
AUTOMATION AND WORK.” Edited by Ajay Agarwal, Avi Goldfarb, and Joshua Gans,
January 2018. https://www.nber.org/system/files/working_papers/w24196/w24196.pdf.
