Recent breakthroughs in the development of artificial intelligence (AI) have triggered profound discussions regarding its implications for the global workforce, emphasising both the potential for innovation and apprehension over widespread job displacement.
Past industrial revolutions have tended to impact jobs (for good and bad) at the lower end of the skills spectrum: the mechanisation of manual labour in the 18th century; the rise of routine jobs in mass-production lines in the 19th century; and the automation and computerisation of manufacturing jobs in the 20th century. However, the fourth industrial revolution, characterised by digital technologies and AI, is complicating the historical narrative: global studies have highlighted its impact on higher-wage earners.
AI Is Projected to Have the Greatest Impact on Knowledge Workers
The International Monetary Fund (IMF) has predicted that almost 40 per cent of global employment will be exposed to AI, with the number climbing to 60 per cent in advanced economies due to their proportion of highly skilled jobs. These predictions are in line with other global studies indicating that “knowledge workers” will be most prone to automation and augmentation. With reference to the emergence of generative AI (GenAI) and the advent of large language models (LLMs) such as ChatGPT, the International Labour Organization (ILO) has predicted that high-skilled, white-collar occupations (such as managers, chief executives, and science and engineering professionals) are likely to face considerable exposure to AI’s recent progress.
These occupations, previously assumed to be safe from job displacement, have recently come under scrutiny due to GenAI’s startling ability to automate the type of tasks that are part of many white-collar jobs: analysing large datasets, information synthesis and content generation, all at astonishing speeds. Addressing the impact of GenAI, industry projections indicate that it could lead to the loss of 300 million full-time jobs, and that half of today’s work tasks will be automated somewhere between 2030 and 2060.
However, caveats remain. Limitations (such as the speed at which AI is being adopted, and bottlenecks due to certain skills being more resistant to automation) challenge the accuracy of these estimates. So too do rapid technological advancements: an update to a decade-old study that estimated up to 47 per cent of jobs in the United States were at risk indicates that, instead, lower-skilled workers actually stand to benefit most from GenAI.
These findings are echoed elsewhere too. A study of customer-support agents found that when call-centre operators used GenAI-based assistants in their work they became 14 per cent more productive. There was also a 34 per cent increase in productivity for the lowest-skilled agents, who moved up the learning curve faster with GenAI assistance. Another study on the use of LLMs, involving more than 750 business consultants, reached a similar conclusion.
There is also agreement that most jobs that are partially exposed to automation are more likely to be transformed than fully automated, emphasising the potential of augmentation in many occupations.
Asia’s Unique Labour-Market Environment Warrants a Closer Look
Although some degree of consensus can be found in studies from around the globe, most of the above insights were derived from research that tapped into the O*NET database. This is a comprehensive, standardised, occupation-specific set of descriptors covering the entire US economy that is generally applicable to many Western economies. However, the data may have limited application in Asia, not least due to the continent’s unique labour-market environment and regional concerns.
In South Asia, large parts of the workforce are in the informal sector, where employees don’t have fixed working hours and wages. As such, their data are not adequately captured. Informal work in Asia rose from an average of 50 per cent in the 2000s to 60 per cent over the period between 2010 and 2016. In India alone about 90 per cent of the workforce engages in informal work, notwithstanding fast economic growth and technology adoption. Data capture aside, there are pertinent concerns over the lower wages and protections for workers in the informal sector, as well as the knock-on effect for productivity.
The growth of the informal sector in Asia can be seen through the increasing prominence of digital labour platforms, which mediate jobs that are short-term, task-based and on-demand. Increasingly, studies have highlighted the emergence of “microtasks” (image and data annotation, labelling, categorisation and data processing, for example) on such platforms. These tasks, which power the development of AI systems in advanced economies, are often outsourced to workers in South Asia and Africa. Studies highlight how some microtasks can be harmful to workers’ mental health (such as reviewing disturbing images or videos) and are often unrelated to their educational qualifications. More importantly, the poor quality of such jobs can lead to the deskilling of workers and the creation of low-end service economies.
While digitalisation can create jobs and raise productivity and competitiveness, there are concerns that GenAI could imminently absorb roles in the business-process outsourcing (BPO) sector in the Philippines. BPO is a key sector: it employs about 1.3 million workers in call centres, knowledge-process outsourcing and back offices, software development, game development and the like, generating $30 billion a year (or 7 per cent of GDP). More than half of employers in information technology BPO in the country – 59 per cent – reported that they had adopted “4IR technologies” (referring to data exchange technologies, cyber-physical systems, the internet of things, AI, cloud computing and cognitive computing).
With the forecast that at least 1.1 million Filipinos will have lost their jobs as a result of AI by 2028, Senator Imee Marcos has filed a resolution to the Senate of the Philippines to review AI use in the country’s factories and call centres. While more jobs in developed markets are exposed to automation than in emerging markets, it is predicted that nearly 20 per cent of jobs in the Philippines – and more than 10 per cent in India – could be lost to AI.
More Efforts Are Needed to Achieve a Balanced Distributional Effect
Historical evidence suggests that technological advancements can boost wages without reducing the labour share. While some jobs will inevitably be lost or transformed, the impact will be counterbalanced by the beneficial productivity effect and the creation of novel job opportunities; this can lead to market expansion and the emergence of new sectors through technological innovation. AI may have the potential to create positive impact on workplace safety but concerns persist regarding job displacement, job quality, income distribution and gendered impact.
Asia’s unique labour-market environment means that low- and middle-income countries in the region need to act urgently to realise the potential of digitally driven jobs as a step towards improved sustainable development. Failure to act promptly could exacerbate digital and wage inequality, and ramp up social tensions. Policymakers need to consider the current and longer-term needs of the workforce and economy, taking into account ways to integrate the future of work with economic and industrial strategies. While new jobs are being created through the development and adoption of new AI technologies, governments need to ensure that their economic, education and skills strategies work in a cohesive way to bring about structural transformation, sustainable growth and good work.
Given the significant presence of BPO jobs in Asia, one focus area could be to encourage and incentivise employers to support middle- and low-skilled workers in benefitting the most from AI technologies, putting the emphasis on tech adoption and upskilling. Education and skills strategies also need to go hand in hand with the needs of the economy to ensure that the skills and talent pipeline is robust. People should be able to not only use and adopt technology, but also innovate and create value-added businesses.
A Multifaceted, Participatory Approach Is Needed
Acknowledging the dual nature of AI’s impact, researchers have proposed policy measures to manage disruption effectively. Existing recommendations include strategic workforce planning, lifelong learning systems and social safety nets. Other integral components of policy suggestions are the enforcement of principles for the use of AI, support for low-wage workers and consultation between employers and workers.
AI’s impact on the future of work will depend on the paths and decisions that policymakers take. The IMF findings revealed that, although advanced economies face greater exposure to AI, they are also more prepared for the disruption that it will bring. Countries such as Singapore and the US, which are top of IMF’s AI Preparedness Index, are poised to take advantage of the AI wave thanks to their consistent and strategic investment in human capital, digital infrastructure, innovation and AI governance. As such, policymakers in other countries should explore ways to enable a systematic approach to developing human capital for the fourth industrial revolution, not least by focusing on linking up education and skills with industrial strategies.
Given the rapid development of AI, it is paramount that multi-stakeholder groups come together – and this includes a voice for workers, who are so often on the receiving end of technology adoption. Countries should put concerted effort into uncovering real-life insights on the ground and take a participatory approach to engage workers, employers and unions alike. The disruptive nature of technology warrants this multifaceted and inclusive approach so that all of us can benefit from its transformational potential.
Lead image: Getty