The impact of AI on the labor market or whether robots should replace people

AI is quickly transforming the world economy and employment landscape. As AI tools improve and become more widely available, they bring both fresh possibilities and major issues for society.

AI's reach in the labor market

A new IMF study shows that AI affects nearly 40% of jobs globally. The degree of impact varies significantly depending on countries' economic development levels:

- In developed economies, about 60% of jobs may be affected by AI.

- In emerging market countries, this figure is around 40%.

- In low-income countries, it's about 26%.

These numbers highlight that developed countries face greater risks but have more opportunities to benefit from AI adoption.

It's worth noting that AI's impact on jobs doesn't mean they'll disappear completely. About half of the jobs affected by AI could benefit from integrating these technologies, increasing labor productivity. For the other half, AI might perform key tasks, potentially leading to reduced labor demand, lower wages, and decreased hiring.

A Goldman Sachs study showed that generative AI could create new jobs and increase global productivity, but it could lead to "significant labor market disruption," with about 300 million jobs at risk of automation.

British Telecom's example demonstrates the reality of these forecasts: the company announced plans to cut up to 55,000 jobs by 2030, with the possibility of replacing 10,000 of them using AI. At the end of 2024, they have a total of 99,000 employees. These numbers are very striking.

Acemoglu and Restrepo proposed a theoretical framework for understanding new technologies' impact on the labor market. They identify three main effects: the displacement effect, the productivity enhancement effect, and the new task creation effect. The latter is particularly important, as new technologies can serve as a platform for creating new tasks in services where labor has a comparative advantage over machines, stimulating labor demand.

Industry and profession transformation

AI's impact on various industries and professions is uneven. Studies show that areas such as marketing, content creation, technical support, and sales will undergo rapid automation. Medium-speed automation can be expected in programming and scientific research, where AI is more likely to become an assistant to specialists. Slow automation will affect regulated fields such as medicine, healthcare, and education due to strict regulations and the need for verification.

Automation of physical labor will depend on infrastructure readiness. For example, self-driving cars may be widely used faster than construction robots or home assistants.

AI may disproportionately and negatively affect socio-economic groups that have historically faced the greatest obstacles in the labor market. Scientists increasingly call for AI analysis that specifically examines gender and racial biases.

The adoption of AI technologies shows considerable variation across different industries. Acemoglu and Restrepo's research indicates that the service sector, particularly advanced fields like information, professional, scientific, and business services, leads in AI implementation.

Some utility sectors, such as electricity, and certain public sector areas, including national security and international relations, also show significant AI adoption. In contrast, the manufacturing sector has seen limited AI integration so far. This pattern differs from the use of industrial robots, which are primarily found in manufacturing.

Economic consequences and productivity

AI adoption promises significant increases in productivity and economic growth. Price Waterhouse Coopers predicts that AI could increase global GDP by 14% by 2030. McKinsey Global Institute forecasts that about 70% of companies will adopt at least one type of AI technology by 2030, and less than half of large companies may use the full range of AI technologies.

AI's impact on overall factor productivity may be moderate. Acemoglu's research shows that the expected productivity increase will be only 0.66%–0.71% over ten years. These benefits are likely to diminish over time, as early AI applications focus on simple tasks, while future ones will include more complex ones.

AI's impact on the economy may be uneven. AI could increase income inequality between capital and labor. Workers with high skill levels who can utilize AI technologies might experience increases in productivity and wages. In contrast, those with lower skill levels could potentially encounter job losses or stagnant wages.

Research conducted in China, analyzing data from 30 provinces between 2006 and 2020, revealed that the implementation of AI technologies, specifically industrial robots, led to an increase in job numbers. This finding challenges the common view of robotics as a factor that reduces employment, suggesting that AI's overall effect on the job market could be beneficial.

The proportion of jobs related to AI in the United States saw a significant increase from 2000 to 2020, rising from 0.14% to 0.20%. A large part of this increase took place in the years following 2010.

Challenges and adaptation paths

Despite potential benefits, AI adoption creates several challenges for society and the economy:

1. Inequality: The gap in income and wealth may widen between people who can use AI effectively and those who cannot.

2. Requalification: Workers will need extensive retraining and skill improvement to meet new job market demands.

3. Regulation: New rules must be created to control AI use and safeguard workers' rights.

4. Ethical issues: We must tackle ethical concerns related to AI, including data privacy and biased algorithms.

5. Country readiness: The IMF developed an index to measure countries' readiness for AI, considering digital infrastructure, economic integration and innovation, human capital level and labor market policy, and regulation and ethics. Among 30 assessed countries, Singapore, the USA, and Germany top the list, while middle-income countries are at the bottom along with low-income countries.

To successfully adapt to the AI era, several measures are necessary:

- Invest in education and retraining programs for workers.

- Develop digital infrastructure, especially in developing countries.

- Create support policies for workers at risk of job loss due to automation.

- Improve the social security system and accelerate the development of high-class domestic robots.

- Deepen the reform of the education and vocational training system.

Some researchers suggest considering the introduction of Universal Basic Income (UBI) as a way to mitigate the negative consequences of automation. There are many forms of basic income, and several dozen experiments have been conducted from the 1970s to the present, involving millions of people. The results of these experiments are consistent: less anxiety, better children's education, more inclination towards entrepreneurship, while the desire to work does not decrease.

UBI implementation faces several problems. One is inflation and often the unpreparedness of the fiscal system. For example, the current US social security budget is $1.1 trillion, while introducing UBI would cost $2.8 trillion. There's a risk of power centralization with state or corporate UBI, though there are potential solutions, including projects like Worldcoin or more decentralized alternatives.

In conclusion, AI's impact on the labor market and economy will be significant but not necessarily catastrophic. With proper management and adaptation, AI can become a powerful tool for increasing productivity and creating new opportunities. The key success factor will be society's and governments' ability to effectively coordinate the introduction of these technologies, ensuring fair distribution of benefits and minimizing negative consequences.

The main challenge with technologies lies not in their invention, but in their implementation and in establishing new ways for society to function. The critical issue is whether our society can adjust to these shifts and apply them to benefit everyone, rather than just a small group.