Gates Foundation launches coalition making AI inclusive for underrepresented languages
The Gates Foundation just kicked off a major coalition, teaming up with 60 partners like Google and Anthropic, to make AI tools work better for people who speak underrepresented languages.
Over the next five years, they're hoping to reach more than 3 billion people worldwide by making language data in AI much more inclusive.
Coalition to build diverse language datasets
Mark Suzman, the foundation's CEO, says it's about unlocking AI's power to help everyone, not just those who speak big global languages.
One big project is Google's Project Vaani, which will collect speech samples from every district in India so local dialects aren't left out of future tech.
The coalition also plans to build diverse language datasets so that AI reflects real-world cultures and voices, not just what's already online.
As Mozilla Data Collective CEO EM Lewis-Jong puts it, AI models need data that actually looks like the world we live in.