Will developing nations be able to handle millions of AI agents?

The challenges of deploying AI agents in developing nations

The full report

Meta's Muse and OpenAI's Dots are personal agents designed to perform digital tasks for users, but their effectiveness in developing nations is uncertain. According to AppMagic, Meta's Muse has been downloaded 9 million times in the U.S. and Canada since its launch on September 8, with as many as 1.8 million people using the app daily.

Meta aspires to a global launch, but adapting agents to local conditions has its challenges. Anton Moulder, founder of South African startup ManyHands, said that getting the AI assistant to converse in all of South Africa's 11 official spoken languages has been difficult.

AI researchers and experts have expressed concerns about the limitations of large language models in non-English languages. Payal Arora, professor of inclusive AI cultures at Utrecht University, said that 'linguistic inclusion is ultimately about power' and that agents might impose Western assumptions on their users.

Agents would also need to understand the local culture, including the unique ways people communicate and bargain. Usha Raman, an independent researcher on AI policy and digital culture, warned that agents might 'impose Western assumptions on their users' and 'force the rest of the world to adapt'.

Another challenge is the potential for agents to stress networks and websites, particularly in regions with poor internet connectivity. Kentaro Toyama, professor of community information at the University of Michigan, noted that 'internet connectivity may be poor [in developing nations], affecting not only the user's interaction with AI, but what agents can do online'.