PublicAI be making AI more human on NEAR

I’ve been digging into projects on NEAR lately, and PublicAI really stood out to me. Most AI-crypto plays feel like buzzwords slapped together, but this one is more about how AI actually gets built and trained.

Instead of AI being trained behind closed doors by big companies, PublicAI makes the process open. People can contribute data, text, audio, even brainwave signals, and get rewarded for it. The cool part is that it isn’t just dumping data blindly; everything is verified through a staking and judging mechanism built on smart contracts. That means the community helps decide which data is useful, and contributors earn $PUBLIC tokens for their work.

It also isn’t locked to one chain. While it’s building on NEAR, it has compatibility with Solana too, and the setup is designed to handle real-time model-assisted labeling. Basically, it’s creating an ecosystem where AI models get better through direct human input, while everyone involved has a chance to benefit economically.

The $PUBLIC token is listed on exchanges like Bitget and others and at the center of it, it’s used for staking, validating, and revenue sharing. What I like is that rewards are tied to actual contributions and usage, not just speculation. If data buyers or AI training partners use the system, value flows back to the people who helped build it.

It feels like a more grounded way of blending AI and Web3. Has anyone here tested PublicAI yet, or followed how it’s evolving within the NEAR ecosystem?