Artificial intelligence is changing the structure of our global economy, but it is unlikely that everyone will benefit. AI advocates celebrate its potential decoding intractable global challenges and even end povertybut its achievements in this regard are meager. Instead, global inequality is now set to increase. Countries where AI is developing and are easily able to integrate these technologies into industry are expected to experience increasing economic growth. But the rest of the world, facing major barriers to AI adoption, will increasingly be left behind.
The introduction of new technologies into society has historically led to economic development and growth. Technologies are often designed for this purpose by increasing productivity: the sewing machine or the tractor, for example, made it possible to manufacture textiles or harvest crops more quickly. Since the turn of the century, digital technologies have been a particularly powerful economic force. In the United States, according to Study 2021the Internet’s contribution to the country’s GDP has increased by 22% annually since 2016. America’s digital economy is now value well over $4 trillion.
Artificial intelligence is changing the structure of our global economy, but it is unlikely that everyone will benefit. AI advocates celebrate its potential decoding intractable global challenges and even end povertybut its achievements in this regard are meager. Instead, global inequality is now set to increase. Countries where AI is developing and are easily able to integrate these technologies into industry are expected to experience increasing economic growth. But the rest of the world, facing major barriers to AI adoption, will increasingly be left behind.
The introduction of new technologies into society has historically led to economic development and growth. Technologies are often designed for this purpose by increasing productivity: the sewing machine or the tractor, for example, made it possible to manufacture textiles or harvest crops more quickly. Since the turn of the century, digital technologies have been a particularly powerful economic force. In the United States, according to Study 2021the Internet’s contribution to the country’s GDP has increased by 22% annually since 2016. America’s digital economy is now value well over $4 trillion.
AI is a powerful new force for economic growth. In 2017, PwC attempted to quantify the value AI would bring to national economies and global GDP. In a founding report entitled “Sizing the price”, the consultancy boasted that by 2030, AI would contribute $15.7 trillion to the global economy. China, North America and Europe are expected to take 84% of this price. The rest is scattered throughout the rest of the world, with 3% planned for Latin America, 6% for developed Asia and 8% for the entire “Africa, Oceania and other Asian markets” bloc, as it is called. PwC.
Following the advent of generative AI technologies such as OpenAI’s GPT series, McKinsey It is estimated that this new generation of AI would increase the productive capacity of AI across all sectors by 15-40%, which could bring up to $4.4 trillion per year to the global economy. These estimates are widely considered conservative. The capabilities of the new suite of large language models, of which ChatGPT is a part, are particularly important for their ability to increase productivity levels, particularly in knowledge economies where language-based tasks form the basis of production.
The McKinsey report also includes a breakdown of the sectors and productive functions likely to see the highest growth, particularly high-tech industries (technology, space exploration, defense), banking and retail. In contrast, the industry likely to see the slowest growth is agriculture, Africa’s most important sector by far and the main source of livelihoods and jobs on the continent.
However, McKinsey’s calculations date back to the beginning of the generative AI revolution, when information on how AI technologies could improve agricultural production in developing contexts was limited. Today, there are a growing number of use cases demonstrating the value of AI in African agribusinesses. In Tanzania, a researcher at the Sokoine University of Agriculture uses generative AI technologies to create a app for local farmers to use to receive advice on crop diseases, yields and local markets to sell their products. In Ghana, experts from Responsible AI Lab are designing AI technologies to detect unsafe foods. Yet cases like these remain limited in scale and impact. At this stage, it is unclear whether AI will be as transformative in African contexts as it promises.
AI adoption in developing regions is also limited by design. AI designed in Silicon Valley using largely English-speaking data is not often fit for purpose outside of rich Western contexts. Productive use of AI requires stable access to the Internet or smartphone technology; in sub-Saharan Africa, only 25 percent of people have reliable Internet access, and it is estimated that African women are 32% less likely to use mobile internet than their male counterparts.
Generative AI technologies are also primarily developed in English, meaning the results they produce for non-Western users and contexts are often unnecessary, inaccurate and biased. Innovators in the Global South must spend at least twice as much effort to make their AI applications work in local contexts, often by retraining their models on localized datasets and employing extensive testing practices and errors.
Where AI is designed to generate profit and entertainment only for the privileged, it will not be effective in combating conditions of poverty and changing the lives of marginalized groups in AI consumer markets. Without a high level of saturation in key industries and without the infrastructure in place to enable meaningful access to AI for all, it is unlikely that countries in the Global South will be able to reap major economic benefits from this technology.
As AI is adopted across industries, human work is evolving. For poorer countries, this is creating a new race to the bottom where machines are cheaper than humans and cheap labor that was once offshored to their lands is now offshored to rich countries. Those most affected are those with lower education levels and fewer skills, whose jobs can be more easily automated. In short, a large part of the population in low- and middle-income countries could be affected, with serious consequences for the lives of millions and threatening the ability of poorer countries to prosper.
Generative AI technologies threaten the rising middle class in developing contexts. A recent report from World Bank estimates that up to 5% of jobs are at risk of being fully automated by generative AI in Latin America and the Caribbean and that women are most likely to be affected. In countries where creating formal jobs and economies is a major development priority, AI is expected to push millions of people into insecure temporary, on-demand or contract employment.
In fact, on-demand economies are growing rapidly. Currently, research estimates that the gig economy’s global market share is $500 billion, but is expected to grow to almost $500 billion. 2 trillion by 2032. Several million gig workers (around 30 to 40 million) come from all over the South. Workers in platform economies, like delivery drivers, often hold multiple jobs in order to earn just enough to survive and certainly not enough to escape a life of poverty. Globally, platform and gig workers have limited labor rights, Global Responsible AI Index finding that only seven countries in the world have enforceable laws protecting these workers.
While AI creates uncertainty for the poor, we are seeing the largest transfer of income to the highest strata of society. Globally, two-thirds of all wealth generated between 2020 and 2022 was amassed by richest 1 percentaccording to Oxfam estimates. And the richest among them is the new class of tech billionaires, with the power, money and influence to create the world they want to live in. Technology companies are some of the largest companies in the world. Apple, which ranks among the world’s five largest companies, has a market capitalization that exceeds the combined total GDP of the African continent.
The wealth of tech companies not only paints a picture of the glaring inequalities at the heart of AI; it also creates a barrier to the production of AI technologies by other actors. Recently, OpenAI CEO Sam Altman embarked on a campaign to increase 7 trillion dollars to power an AI-driven future. That’s the kind of money needed to establish the kind of supercomputing infrastructure needed to create cutting-edge AI models. It’s not a sport that everyone can afford.
Compute, essential for creating AI technologies and applications, is one of the most expensive resources in the world. There is a major global gap in access to computing resources. Collectively, the Global South is home to just over 1 percent of the world’s best computers, and Africa only 0.04 percent. Right now, with the kind of computing capacity available in Africa or South America, it would take hundreds of years to catch up with the progress made with generative AI in the West and developed countries in the East.
The costs for poorer countries to catch up in the AI race are too high. Public spending can be diverted from essential services such as education and health care. While governments in the Global South should be attuned to the AI revolution, policymakers should closely evaluate the effects of AI on their economies.
