‘Small AI’ Could Transform Developing Countries’ Economies; World Bank Calls for Technology Built for Mobile Phones and Local Languages

The World Bank has called for building artificial intelligence (AI) technology that can be used on mobile phones, works effectively even with low internet connectivity, and is based on local languages, to accelerate the economic progress of developing countries. According to the organisation, using ‘small AI’ could make it possible to cover a centuries-long development path within a few decades.

Pascale Donohoe, the World Bank’s Chief Knowledge Officer, said that while big technology companies focus on creating advanced consumer products, it is essential to build AI technology suited to the needs of developing countries. In her view, if technology is not built in line with real needs, its benefits will remain stuck within a limited sphere.

Specialised models trained on limited datasets for specific tasks are called ‘small AI’. It can be used in local languages, on ordinary mobile phones and even with weak internet connections — these features are what set it apart from big and expensive models and make it suitable for reaching users at the margins.

As an example, Donohoe mentioned India’s farmers. Under the World Bank’s agricultural development initiative ‘AgriConnect’, farmers need information on soil quality analysis in their own language. Since it is difficult to ensure advice from agricultural experts for every small farm, AI can give farmers the information and guidance they need.

She said such technology can be built through joint initiatives of the World Bank, governments and the private sector. That is, rather than depending only on technology companies, solutions based on local needs should be built through coordinated efforts of development partners.

At the annual meetings of the World Bank and the International Monetary Fund (IMF) held in Bangkok, the impact of AI on employment and growth was discussed. It was there that these calls and analyses came to the fore.

IMF Managing Director Kristalina Georgieva said that to gain the benefits of AI, it is essential to ensure electricity, digital infrastructure, readiness to use technology, and safety. Her remarks also suggested that technology alone, without infrastructure, will not bring benefits.

The World Bank’s World Development Report has also brought out a picture of infrastructure disparity. It says nearly one-third of rural schools in sub-Saharan Africa have no electricity, and more than two-thirds of the schools lack reliable internet connections. These deficits are major obstacles on the path to delivering AI-based education and services.

According to Georgieva’s warning, around 40 percent of jobs worldwide could be changed or eliminated by the impact of AI. Young people entering working life are at particular risk. As a result, alongside the opportunities of technology, preparation to handle the risks to employment has also become essential.

Meanwhile, the World Bank has said that over the past year it has helped mobilise 120 billion dollars of private capital through guarantees and other financial arrangements. At the meeting it was also discussed that technology-sector initiatives could be linked to this effort to draw private capital into development financing.

All in all, the message is clear: if small, affordable AI that runs in local languages and on weak connections can be built, going beyond the competition of big models, rapid change can be brought to the agriculture, education and service sectors. But if deficits in basic infrastructure such as electricity and the internet are not met, it will be difficult for that potential to become reality.

How governments, development agencies and the private sector join this ‘small AI’ initiative in the coming days, and how far it reaches users at the margins such as farmers, will be the real test of its success.