The increasingly rapid adoption and routine use of artificial intelligence (AI) applications could increase efficiency across multiple economic sectors, manufacturing, and business practices, according to a new study by TRENDS Research and Advisory.
"Improvements in computational power and the growth of machine learning have also emerged as key drivers underpinning the growth of AI. The use of machine learning by AI allows relatively accurate prediction of outcomes by training algorithms to perform predictive tasks using historical data as an input, without the need for human programming intervention,’’ said the study, authored by Nouf Yaqoob Al-Saadi, a Research Assistant. The growth of big data is another factor underpinning the rapid adoption of AI and machine learning in everyday life. With over 4.95 billion internet users globally, constantly expanding data sets enable the development of evermore accurate algorithms and applications.
Examples of AI in daily life include smart assistants (Apple’s Siri and Amazon’s Alexa), chatbots, predictive mapping apps, personalised e-shopping, AI-driven hiring, algorithmic management of taxi services, facial recognition software, autonomous vehicles, threat prevention in data security, and many other applications.
However, the speed at which AI and machine learning are developing have focused on several potential policy problems, including worker displacement and labour market disruptions. AI-powered systems can perform tasks previously done by humans, leading to the displacement of certain jobs. Moreover, given the fundamental role of labour as one of the four factors of production for goods and services, the dislocation of labour markets and displacement of workers could result in broad-based effects across employment levels, wages, and labour market opportunities.
Early analyses of the AI phenomenon focused on the potential of the technology to replace large swathes of jobs globally, with a well-known management consulting study from 2017 estimating that AI and automation could effectively displace between 400 and 800 million jobs by 2030. Researchers from Stanford University have also observed that, by its nature, AI “performs tasks that involve detecting patterns, making judgments, and optimising. The most-exposed occupations include clinical laboratory technicians, chemical engineers, optometrists, and power plant operators.”
More recently, the OECD concluded that the “direct displacement” effect of AI exposure on workers is not statistically significant but observed that workers with highly developed digital skills likely “find it easier to use AI effectively and shift to non-automatable, higher-value-added tasks within their occupations.” Conversely, the prospects of workers with lower-level digital skills are hindered by their inability to interact efficiently with the technology and avail of its potential benefits.
This perspective advocates that artificial intelligence and machine learning are not replacing workers in the traditional sense; rather, the effect of the technologies is to make adjustments to the labour market system as a whole. As an analogy, while “digital computers” have transformed the nature of work in almost every sector of the economy over the past several decades, the net result has been accelerating productivity and complementary innovations rather than outright replacement of workers.
Focusing on industries that seem to be most affected by AI and machine learning, the marketing industry case illustrates an example of complementary innovation. Large-scale marketing campaigns, traditionally essential to reach targeted audiences or potential consumers, have been complemented by the ability of machine learning to access and learn from data that enable accurate predictions of the interests of consumers. In this way, machine learning makes it possible for advertisements to be tailored more accurately to reach the maximum number of potential consumers.
Healthcare is another industry that is currently being impacted by machine learning. An article published by the World Economic Forum in 2020 concludes that by 2030, AI will access multiple sources of data to reveal patterns in disease and that the technology will enable predictions of an individual’s risk of certain diseases while also suggesting preventative measures.
Beyond medical diagnoses, AI is also expected to contribute to administrative efficiencies in the healthcare system by helping to reduce waiting times for patients. Because of mandatory procedures such as Electronic Medical Records (EMR), healthcare systems have already used big data tools for next-generation data analytics. Machine learning tools are poised to add even more value to this process.
When we speak of education, we think of schools, institutions/universities, teaching, and knowledge access. For years, computers have been widely used for various educational processes. However, the advent of machine learning is fundamentally transforming teaching, learning, and research methodologies.
Through adaptive learning, for example, this technological innovation can evaluate student performance in real-time and modify teaching strategies in accordance with the findings, providing students with individually customised content based on their needs. This is particularly useful for detecting students with specific learning difficulties and improving their performance and retention in class. Another positive outcome of machine learning is the higher efficiency in managing schedules and classroom content, which largely frees teachers from administrative tasks, allowing them more time to concentrate on tasks requiring human interaction and creative approaches.
Similarly, customer service representatives using AI to automate routine interactions with customers would free up their time to focus on more intricate customer issues. By automating repetitive tasks, workers can allocate their energy and attention to more intricate and fulfilling work.
The examples mentioned above illustrate how the use of AI can create new opportunities for workers to learn new skills and take on higher-value jobs. Additionally, AI and machine learning are creating a new trend in “desirable” skills on a global scale, enhancing the usual skills requirement in jobs. For example, a recent study by the World Economic Forum projects that by 2025, AI will automate 75 million jobs worldwide, generating 133 million new jobs.
While it is likely true that specific roles may become redundant, current trends across these industries suggest that the focus will shift towards upskilling and reskilling workers to adapt to new roles, which will require a mix of technical and soft skills. For these reasons, the main impact of AI on the global job market is likely to be the creation of large-scale demand for reskilling and upskilling.
Supporting this growing trend is the recent proliferation of AI-related job titles, which include: AI Research Scientist, AI Data Analyst, Machine Learning Engineer, Deep Learning Engineer, AI Content Editor, AI Chatbot Strategist, AI Products Content Specialist, and AI Prompt Manager, among others. The economic growth potential of AI is considerable, with management consulting firm PWC estimating that AI, robotics and smart automation could contribute US$15 trillion to global GDP by 2030.
The scope of this anticipated transformation in the global job market poses challenges and opportunities for both employees and employers. In practical terms, it will be crucial to make significant investments in training programmes to upskill and reskill workers in response to the new demands of integrating existing jobs with AI and for new jobs due to the growing demand for workers with skills and experience related to AI. For such training programmes to be effective, continuous research and studies are essential to keep pace with the changing landscape of the labour market.
In conclusion, machine learning is transforming the labour market and will continue to do so, as it is rapidly becoming an established factor in accelerating automation and improving work efficiency. However, the influence of AI on the labour market is subject to the extent and pace at which it is adopted and integrated into workplaces, which still needs to be determined on a global scale.
However, an observation is necessary at this stage: given that the profit motive of private firms could skew investment in AI towards pursuing increasing automation of tasks (as opposed to upskilling existing workers), due consideration must be paid to direct adequate amounts of public funds towards AI research that favours AI as a complement to workers instead of an outright replacement.
It is crucial to approach the integration of AI with current labour markets with an open mind, embracing the potential it has to offer while also being mindful of its limitations and implications for the workforce.