AI for All of Us

A core tenet of DWeb is recognizing that everyone, no matter where they are in the world, should be able to learn, use, and build with technology, especially powerful tools like AI. Today, however, this isn’t the case. While some cultures and communities are well-represented in public web data, many people’s voices and values are not. In order to make AI work for all of us, we must build pathways for people to ensure their languages, cultures, and experiences are included.

How do we ensure that AIs train on data that are representative of the world - linguistically, culturally, and academically? What will it take to create public open data archives that are truly reflective of the world? We hear from the Open AI lead whose goal is to diversify training data, archivists, and indigenous language activists who are all working from different directions to create troves of data that are open, accessible, and inclusive.