Nov 5, 2024
Artificial intelligence is an emerging market, and like all emerging markets, there is lots of uncertainty and many possible outcomes it could have in other markets and ecosystems. The most discussed one of these markets is the labor market.
Since AI has the potential to replace many traditional workers through automation with lower costs, the concern for greater income inequality is on the rise. Deriving from the data, it is safe to suggest that these concerns are for numerous good reasons.
Over the past years, in the rising period of AI, in some industries AI has proven to be more efficient than human workers. A great example of such an industry could be the customer service industry.
Chatbots have been around for a couple of years, but with the latest developments regarding LLMs in the AI sector, chatbots have significantly improved their communication and aiding skills. Day by day, companies increase their use of chatbots, which decreases the need for a human customer service agent.
Even though customers mostly prefer a human customer service agent to a chatbot, the increased usage of chatbots by companies and their developed capabilities create an expectation for this preference to change.
The U.S. Bureau of Labor Statistics foresees a 5% decline in employment for this particular sector until 2033[^1].
Customer agent services is not the only industry where AI has been replacing human workers. Most of the lower-wage routine jobs are vulnerable to AI automation.[^2]
A new report published by McKinsey in May 2024 proposes that up to 30% of hours worked could be automated by generative AI by 2030[^3]. They also state that, according to their research, this rate would cause a number of occupational transitions from 8.5 million to 12.0 million, which accounts for 4.6% to 6.5% of current employment in Europe.
It could be suggested that the automated work hours would account for a significant amount of the occupational transitions which will occur during this period.
In the same report, McKinsey foresees a decrease in total hours worked for both basic and higher cognitive skill-required tasks in the US and Europe. On the other hand, they foresee an increase in harder-to-automate skilled tasks such as social and emotional tasks and technological tasks.
Based on these predictions, it is safe to argue that most of the burden of displacement will fall on less-educated, lower-wage workers. Conversely, the demand for non-automatable skills will be on the rise for the next ten years.
A survey McKinsey conducted among C-suite executives in five countries shows that even today, firms struggle to find employees skilled in areas AI fails to automate. The foreseen increase in demand for such employees will make these assets even more valuable in the next decade.
The decrease in demand for less-educated lower-wage workers and their replaceability by AI will have a measurable impact on the bargaining power of unions, resulting in lower income for this large fraction of the population compared to the more-educated fraction.
Let alone the decrease in wages, many will have a hard time finding a job with their current skill set, since AI can offer the same skill set to employers at lower variable costs.
On the contrary, the already short supply of high-skilled, more-educated workers will become even more valuable as demand for such employees increases. Firms will compete for such workers, driving up their wages. With the cost savings from AI-based automation, firms will also have greater budgets to spend on these workers.
In addition, capital owners will be less dependent on labor providers due to automated jobs through AI, which will strengthen their bargaining power and create an even more favorable environment to maximize profits.
Capital owners will be the ones to benefit the most from AI, since AI is just another potential labor provider—one without unions, survival needs, or self-awareness. These factors combined make AI a perfect labor provider for capital owners seeking to maximize profit, while still providing a sufficient amount of capital for the lower-wage population to consume goods and services.
The barriers of transition from a lower-wage worker to a high-skilled worker are also high. Limited access to high-quality education, the harder adaptability older workers face, and the high costs of education make it difficult to change one’s profession according to the market’s demands.
These barriers contribute to increasing inequality among the labor force, since only the few with sufficient capital and time can overcome these barriers to benefit from the increasing demand for their labor.
The Nobel Prize-winning economist Prof. Dr. Daron Acemoğlu states:
“While it is possible for generative AI to be a useful tool for non-college workers, I do not see the current path achieving this, and the existing evidence is more consistent with most of the burden of displacement falling on less educated, lower-wage workers.”[^4]
While it is still early to make certain assumptions for the future of AI in general, there are strong arguments and evidence suggesting that use of artificial intelligence will have an impact in increasing income inequality.
Sources
[^1]: U.S. Bureau of Labor Statistics – Customer Service Employment Outlook
[^2]: Nedelkoska, L. and G. Quintini (2018), “Automation, skills use and training”, OECD Social, Employment and Migration Working Papers, No. 202, OECD Publishing
