The looming challenges of wider AI adoption are causing some anxiety about unemployment and severe job stratification. But is it really as dire as it seems? Let’s recall history. Similar concerns were raised when automobiles replaced manual labor, tractors revolutionized field plowing, steam engines transformed metal production, and weaving machines revolutionized fabric manufacturing. Humanity quickly adapted, and certain excesses were swiftly addressed. However, the current scenario is somewhat different. We’re not talking about replacing manual labour with machines or the strength of domesticated animals; we’re talking about replacing human minds with artificial intelligence. It’s nearly the last bastion of humanity, and the first warning bell rings – the prospect of unemployment for many employed sectors.
The current pressing question about artificial intelligence revolves around how powerful unemployment will be in the case of widespread AI adoption. The answers to this question are diametrically opposed.
Technological unemployment or a shortened workweek for the same pay? The dismissal of drivers, call center employees, and bank clerks, or the emergence of new professions needed for the efficient use of AI? There are already cases of victories for workers. For instance, the American actors’ union triumphed in a dispute with film studios over the right to receive money in the event of reproducing images of living actors created using artificial intelligence.
Most likely, this is the first such case, and there could be others following suit. Perhaps this provides grounds to consider the conclusions published in November reports by the Autonomy think tank quite realistic. They suggest that the widespread implementation of AI will result not in unemployment but in a reduced workweek while maintaining the current pay level. Efficiency should also be preserved. Similar trends are likely to be present in EU countries, of course, with some variations based on local peculiarities.
But, as always, the devil is in the details.
Two Scenarios for Future Expansion
Autonomy’s forecasts are based on a rather bold assumption that the deployment of artificial intelligence in the pursuit of increased labor productivity will occur with considerations for principles of fairness. This assumption is bold because fairness is not a characteristic commonly associated with markets, and achieving fairness in the labor market may only be possible through robust public regulation. However, we also remember that overly stringent regulations can severely impede innovation, and AI is undoubtedly an innovation.
The first of the scenarios explored by Autonomy assumes that the increased productivity of artificial intelligence will lead to a 20% reduction in the working hours, while worker salaries remain unchanged. This is expected to occur over the next decade.
The second scenario envisions that AI can boost labor productivity by at least 10%, allowing for a 10% reduction in the weekly work schedule while maintaining the same level of pay.
For the UK, the research indicates that by 2033, a four-day workweek (32 working hours per week) could be achieved within ten years for 8.8 million workers, constituting 28% of the workforce. Approximately 27.9 million workers could reduce their working day by at least 10% if large language models of AI were introduced and used as a basis for increased productivity. This accounts for 88% of the workforce in Great Britain.
Autonomy’s second research also appears to cover the United States, with quite similar results. By 2033, a four-day workweek (32 hours per week) could be achieved for 35 million workers, representing 28% of the US workforce. A 10% reduction in the working day could be realized for at least 71% of workers in the US labor market. This would happen if large language models (LLM) of AI were introduced in workplaces and used as a basis for increasing labor productivity. In total, this accounts for 128 million workers by 2033.
So, does a radiant future lie ahead with a four-day workweek and maintaining the previous salary level? What could possibly hinder such happiness?
AI, You’re Not My Boss
Artificial Intelligence becomes an independent force in the economy, occasionally displacing humans? Three prominent issues come to the forefront: workers’ rights protection, corporate accountability for AI-driven decisions, and the role of the government in regulating it all.
The rapid integration of artificial intelligence across various economic sectors poses a real threat of technological unemployment. A multitude of people could find themselves redundant, not just the least qualified professionals. Even doctors and financial traders, individuals currently proud of their pay levels and performance rewards, are not immune.
The full-scale question arises about compensating displaced workers, retraining them, and adapting to new challenges. Companies, the employers of these displaced workers, may not be keen on creating a warm bath for them. After all, why pay more? Even with government and union pressure, employers might not be charitable to workers they no longer need due to AI. The significant reason lies in global competition. Generous compensation and maintaining the same pay while reducing working hours inevitably increase costs and stifle competitiveness.
The next problem concerns corporate responsibility for decisions and actions implemented through artificial intelligence. Who takes responsibility for a misdiagnosis or flawed financial advice? How does a legal firm answer for errors caused by seemingly correct AI work? Who bears responsibility for accidents caused by AI-controlled taxis or trucks? Liability insurance won’t solve this problem; instead, it will shift risks from AI users to the clients and consumers served by AI. But these risks won’t disappear.
Humanity, business leaders, politicians, and scientists currently lack effective, evidence-based solutions to these problems. We are facing something unprecedented, and the risks are incredibly high.
For example, policymakers are revisiting the idea of a Universal Basic Income (UBI) to protect displaced workers and support economic stability. However, funding such income remains a considerable challenge, with horrific economic and political risks. UBI significantly amplifies the influence of politicians and state institutions, as UBI payments would flow through the government. This intensifies the conflict of interest, as politicians become interested in increasing the number of UBI recipients dependent on them. On the other hand, increased tax burdens inevitably suppress innovative activity, a consequence of needing more funds in government budgets for UBI payments.
We haven’t even begun to discuss the issue of legal responsibility for the consequences of poor decisions generated by artificial intelligence. If an AI-driven taxi takes an inefficient route or a chatbot from a local government fails to provide assistance, what then?
Is there any positive news in all of this?
Perhaps. The explosive growth of AI’s role in the economy hasn’t materialized yet. Frankly, the power of AI significantly outpaces the impact of its application in the economy. Moreover, the negative impact on the job market lags behind the pace of AI proliferation.
Indirectly supporting this, the European Central Bank released controversial findings on November 28, 2023.
The results of this study appear quite contradictory. The implementation of powerful AI tools seemingly led to increased employment in sectors where AI has the most significant impact. It seems that the growth of relevant industries, businesses, the introduction of new products, and services relying on AI use precedes job cuts.
Researchers from the ECB honestly warn that there is no basis for definitive conclusions regarding significant unemployment due to AI’s influence.
Source: The Gaze