PANews reported on August 27 that the decentralized AI data infrastructure Poseidon announced the upcoming launch of the V1 version of its audio data collection application. The application aims to solve the training bottlenecks of AI models in accents, noise, and real-world interactions by collecting high-quality voice data from users around the world.

Users can earn Poseidon Points by uploading voice data. Rewards will be awarded based on accuracy, clarity, uniqueness, and completeness, with additional rewards available for certain tasks. All data will undergo technical standards review, AI content detection, and duplicate screening, and will be recorded and uploaded to Story to ensure security and authenticity.

Poseidon emphasizes that only high-quality data will earn points; fraudulent or low-quality submissions will be blocked. User-submitted data will not simply "disappear" into the database; it will power real-world applications—from phones that understand accents, to smart cars that respond safely on busy streets, to voice assistants that truly understand user needs. Poseidon was incubated by Story and has completed a $15 million seed round led by a16z Crypto.