On July 2nd, Masa, a decentralized AI data chain, announced the network technology architecture and product application scenarios based on FairAI, the core components of which include decentralized AI data infrastructure, fair artificial intelligence attribution and reward mechanism, dedicated data, open-source LLM, and open network. The entire network is composed of three main participants: worker nodes (users), validators, and developers (Oracle nodes), and popular worker nodes have covered X, Web sites, Discord, Telegram, audio-to-text, vector retrieval, LLM, and other data dimensions, dedicated to promoting the construction of a collaborative ecosystem that supports the development of FairAI. At present, Masa has allowed developers to vectorize the audio data of any podcast, and then use the Masa protocol's speech-to-text algorithm for automated processing, separating the speaker and timestamp, and then efficiently processing text information for training AI agents in health and wellness scenarios. At the same time, Masa also allows developers to aggregate product feedback and community sentiment based on conversation messages and user interaction data extracted from the Discord server to provide reference product feedback and so on.