Up-to-date technology brings with it both the sweet hope of greater prosperity and the cruel fear of loss. Satya Nadella, the head of Microsoft, says he is haunted by the fact that the industrial revolution left India, the country of his birth. (Indian manufacturers were hardly on a level playing field – Britain was both their rival and ruler at the time). Many technologies, such as online education courses, have created more buzz than economic growth in emerging countries. Some people worry about this generative thing artificial intelligence (AI) will also disappoint the global south. The biggest winners so far appear to be a group of Western early adopters, as well as San Francisco startups and the “magnificent seven” US tech companies that include Microsoft, which have collectively added an astonishing $4.6 trillion to their market value since Chatgpt launch time in November 2022.
However, artificial intelligence can also change lives in emerging countries. As it spreads, this technology could raise productivity and reduce human capital gaps faster than many solutions before it. People in developing countries do not have to be passive recipients of artificial intelligence, but they can shape it to suit their own needs. Most excitingly, it could support income levels catch up with those in the luxurious world.
The promise of artificial intelligence in developing countries is tempting. Just like in the West, it will be a useful, universal tool for consumers and employees, making it easier to obtain and interpret information. Some jobs will disappear, but fresh ones will be created. Because there are fewer knowledge workers in emerging countries, there may be less disruption and profits for existing companies than in the West. The IMF says that between one-fifth and one-quarter of workers there are most vulnerable to replacement, compared with one-third in luxurious countries.
However, better and more accessible public services can have potentially transformative benefits. Developing economies have long been hampered by a lack of educated, vigorous workers. Primary school teachers in India have twice as many students as their American counterparts, but they are ill-prepared for the fight. Doctors are few in Africa; fewer are properly trained. Entire generations of children are growing up in poorly educated schools, in impoverished health and unable to fulfill their potential in an increasingly global labor market.
As our briefing shows this week, policymakers and entrepreneurs around the world are exploring how artificial intelligence can support. India is combining vast language models with speech recognition software to enable illiterate farmers to ask bots questions on how to apply for government loans. Students in Kenya will soon ask a chatbot questions about their homework, and the chatbot will correct and improve its lessons in response. Scientists in Brazil are testing medical artificial intelligence that helps untrained primary care workers treat patients. Medical data collected around the world and fed into artificial intelligence can support improve diagnoses. If AI can make people in poorer countries healthier and better educated, over time it should also support them catch up with the luxurious world.
Interestingly, these benefits may spread faster than earlier waves of technology. It took over 50 years for fresh technologies invented in the early 20th century to reach most countries. In turn, artificial intelligence will spread through a gadget that many people in emerging countries already have and will soon have: the phone in their pocket. Over time, providing and acquiring chatbots will become much cheaper.
Moreover, the technology can be adapted to local needs. So far, there is little evidence of a winner-take-all approach to artificial intelligence that has benefited US social media and search engine companies. This means that different approaches can work. Some developers in India are already taking Western models and adapting them to local data to provide a swift language translation service, avoiding the high capital costs of building the model.
Another idea that is also gaining popularity in the West is to build your own smaller and cheaper models. A narrower set of capabilities, rather than the ability to obtain every piece of information under the sun, may fully meet specific needs. It’s unlikely that medical AI will need to generate entertaining limericks in the style of William Shakespeare, because ChatGPT does it successfully. This still requires computing power and customized data sets. However, it can support adapt AI in more diverse and useful ways.
Some countries are already using artificial intelligence. The Chinese are second only to America thanks to their technological know-how and the deep pockets of internet giants. The Indian outsourcing industry may be disrupted as some back-office tasks are taken over by generative AI. But it is home to a dynamic startup scene, as well as millions of tech creators and a government eager to exploit artificial intelligence to improve its digital infrastructure. This makes it well-prepared to innovate and adapt. Gulf countries like the United Arab Emirates and Saudi Arabia are determined to build a post-oil AI industry. They already have capital and import talent.
Each country will shape technology in its own way. China’s chatbots have been trained not to broach the subject of Xi Jinping; Indian developers focus on lowering language barriers; The Persian Gulf is building a model for Arabic in a substantial language. While the global south won’t take America’s crown, it could benefit widely from all this expertise.
Teaching aid
Of course, a lot can go wrong. Technology is still developing. Computing power may become too costly; local data will need to be collected and stored. Some practitioners may lack the ability to exploit the knowledge at their fingertips or the motivation to try fresh things. While countries in sub-Saharan Africa have the most to gain from improving human capital and government services, technology will spread more slowly there than elsewhere, without better connectivity, governance and regulation.
The good news is that investments to accelerate the spread of artificial intelligence will be richly rewarded. Much about the AI revolution is still uncertain, but there is no doubt that the technology will have many applications and that it will only get better. Emerging countries have already experienced disappointments. This time they have a great opportunity and the power to take advantage of it.
© 2023, The Economist Newspaper Constrained. All rights reserved.
From The Economist, published under license. Original content can be found at www.economist.com
