Fixing AI’s Blind Spots with a Global Community
AI is only as fair as the data it sees. On the HumanAIx Roundtable, Trish Rosal explained how missing context from emerging regions reinforces bias — and how YGG’s global community fills those gaps.
The AI industry faces many challenges, and one of them is how it handles diversity in the data it learns from. If AI is trained on inputs from just a few places or perspectives, its output will reflect that narrow view and become biased. In response, organizations working at the intersection of Web3 and AI — such as OORT, Holo, and YGG — formed HumanAIx to decentralize how AI platforms collect data and offer participants part ownership of the network. As part of this effort, they are sourcing inputs and data from communities in places like the Philippines, ensuring that voices often overlooked in traditional AI development are included.
Recently, YGG Head of Future of Work Trish Rosal joined the HumanAIx Roundtable hosted by OORT MENA Lead Markose Chenthitta. OORT Foundation Chairman Michael Robinson, Holo Director of Product Alastair Ong, and Neo Business Development Manager Tony Sun also participated to discuss how HumanAIx can help the AI industry overcome concerns on accessibility, privacy, and partiality.
The following is an excerpt from the HumanAIx Roundtable, where Trish describes how AI aligns with YGG’s mission to create opportunities for its community through emerging tech. By helping communities contribute their own data and context, Trish also explains how AI becomes another tool for economic inclusion and grassroots participation — especially in regions where traditional work opportunities are limited.
HumanAIx Roundtable: Web3 Infra Meets AI
Markose (33:46): I know you guys are focusing on gaming and data contribution. You guys are also big on community building and adoption. The question I have is, how do you see gamers and gaming guilds contribute to decentralized AI development?
Trish (34:09): We started in gaming and became known for having guilds, which are basically groups of gamers who help each other learn and level up in the game. But guess what? That's a transferable strength. Because our community became known and grew to be a global force, Web3 AI companies noticed. They needed diverse global human input and started approaching us.
Many of these companies were actually gamifying data collection and labeling, so it became a natural opportunity for us. Even in the context of our games, we were already having our community members grow the skills and progress to jobs or other earning opportunities in the game ecosystems, such as being content creators, community moderators, or esports players.
We also have Metaversity, which is our education arm, teaching skills that help our community members in the gaming ecosystem. AI just became another vertical, another industry skill that we started teaching and exposing our community members to. Since we're all over the globe, it fits very well with the decentralized model.
You need inputs from everyone. What does a chair look like from one country to another? Maybe it's different materials. What trees exist in this country versus another? If no one's giving input from their context, especially in the emerging countries, then AI is not going to learn about these contexts, and it's going to be biased.
You can listen to the full recording on X and follow Trish, Markose, Michael, Alastair, and Matthew to learn more about how Web3 can supercharge the AI industry.
Join the YGG Discord and follow YGG on X for future updates.