The medieval advancements in agricultural technology failed to uplift Europe’s peasants from poverty primarily because their rulers diverted the wealth generated by increased productivity into constructing cathedrals instead.

Economists caution that a similar scenario could unfold with the advent of artificial intelligence (AI) if its benefits end up being concentrated in the hands of a privileged few rather than benefiting the broader population.

Professor Simon Johnson, an expert in global economics and management at MIT Sloan School of Management, highlights the dual potential of AI – it could either lead to prosperity or exacerbate existing inequalities.

AI proponents project a significant boost in productivity that will create wealth and enhance living standards.

McKinsey, a consultancy firm, estimated in June that AI could contribute between $14 trillion and $22 trillion annually to the global economy, with the upper figure nearly equivalent to the entire current U.S. economy.

Some optimistic voices even go so far as to claim that AI, along with robots, will liberate humanity from mundane tasks, ushering us into a new era of creativity and leisure.

However, concerns abound about the implications of AI on jobs and livelihoods. There is historical evidence that technological advances have often resulted in uncertain, uneven, and sometimes detrimental economic impacts.

A recent book authored by Johnson and fellow economist Daron Acemoglu examines a thousand years of technological progress, starting from the plough to automated self-checkout systems, and assesses their success in job creation and wealth distribution.

For instance, while the spinning jenny played a vital role in automating the textiles industry in the 18th century, it also led to longer work hours in harsh conditions for the laborers.

Similarly, the mechanical cotton gins in the 19th century contributed to the expansion of slavery in the American South.

The Internet, though transformative, has also demonstrated mixed outcomes. While it has generated many new job opportunities, a significant portion of the wealth it created has been amassed by a handful of billionaires.

Additionally, the initial productivity gains from the Internet have tapered off in many economies.

Caution is warranted when assessing the potential impact of AI on labor productivity. French bank Natixis cautioned in a research note that AI might not significantly impact some sectors while creating low-skilled jobs, such as those in the delivery chain for online purchases.

In a globalized economy, there are additional concerns about whether the benefits of AI will be distributed evenly. On one hand, there is a risk of a “race to the bottom” as governments vie for AI investments through lenient regulations.

On the other hand, the barriers to entry for attracting such investments may be too high for many developing countries to overcome.

According to Stefano Scarpetta, the Director of Employment, Labour, and Social Affairs at the OECD, the right infrastructure, including substantial computing capacity, is essential for participating in the AI revolution.

Scarpetta advocates for an expanded accord among G20 and UN nations, building on the efforts of the G7 Hiroshima Process, to jointly explore the opportunities and challenges posed by generative AI.

Innovation and technological progress are only part of the equation; the more challenging task is ensuring that these advancements benefit everyone – and that’s where politics comes into play.

Drawing from historical examples, Professor Johnson argues that technological advances can be more inclusive when they coincide with periods of democratic reform.

For instance, the arrival of railways in 19th century England, alongside democratic gains, allowed the benefits of this innovation to be shared more broadly, improving transportation and leisure opportunities.

However, the last four decades have seen a shift towards aggressive shareholder capitalism, which has diminished the widespread distribution of benefits from technological advancements.

The automated self-checkout system serves as an example of this trend, where cost savings from reduced labor expenses translate into increased profits without benefiting consumers or workers.

Worker groups, whose influence has waned since the 1980s, are concerned about AI’s potential threat to workers’ rights and job security.

For instance, if AI-driven decisions govern hiring and firing without human oversight, it could exacerbate job insecurity.

Mary Towers, an employment rights policy officer at Britain’s Trades Union Congress, emphasizes the importance of unions having statutory consultation rights and the ability to collectively bargain regarding workplace technology.

Multiple factors will determine how AI shapes the economy, including antitrust policies to promote fair competition among AI suppliers and workforce retraining initiatives.

An OECD survey of over 5,300 workers revealed that AI could enhance job satisfaction, health, and wages, but it also raised concerns about privacy, workplace biases, and potential overwork.

In conclusion, the impact of AI on society is at a critical juncture. Its potential to uplift living standards and boost productivity is undeniable, but there are also legitimate concerns about job displacement and inequality.

Learning from history, it is evident that the way AI is integrated into society will be a result of political choices.

Ensuring that the benefits of AI are distributed fairly and inclusively will require thoughtful policies and concerted efforts from governments, businesses, and worker representatives.

The question remains whether AI will exacerbate existing disparities or pave the way toward a more equitable future.