World Bank proposes 'small AI' for developing countries
The World Bank is encouraging the development of specialized 'small AI' models capable of running on low-bandwidth connections and mobile devices to accelerate growth in developing economies.
At the annual World Bank and IMF meetings held in Bangkok, it was emphasized that 'small AI' models focused on local needs are critical for developing countries to benefit from artificial intelligence.
World Bank Chief Information Officer Paschal Donohoe stated that if artificial intelligence is adapted to local needs, it could accelerate development in emerging economies. Projects like AgriConnect in India demonstrate that these models, trained on narrow datasets, provide support to farmers in their own languages on topics such as soil analysis.
IMF Managing Director Kristalina Georgieva warned that artificial intelligence could affect or transform 40% of jobs worldwide, stating that governments must provide electricity, digital infrastructure, and security measures for the transition to this technology. Under the leadership of World Bank President Ajay Banga, the institution provided $120 billion in private sector financing last year to support economic development.