Current Problems in the AI Industry
- Centralization: AI models are trained on centralized servers, leading to data privacy concerns and a lack of transparency.
- High Computation Barrier to Entry: AI research is expensive and requires specialized hardware, limiting access to a small group of researchers.
- Lack of Incentives for Researchers: Researchers are not rewarded for their work, leading to a lack of innovation and progress in the field.
- Difficult collaboration: It is difficult for researchers to collaborate and share their work with others.
Description
Our platform leverages a decentralized AI infrastructure built on EigenLayer's restaking mechanism, which taps into Ethereum's security to ensure robust operation for our AI agents. Our AI marketplace fosters innovation by providing access to pre-trained/finetuned models, using a subscription-based royalty system to reward AI/ML researchers. These models can be used by any apps (stock or user specified), powering their AI dApps with a decentralised AI infrastructure. The platform is built on Near Protocol, utilizing its yield resume architecture for high-performance, asynchronous contracts in Rust. This combination of technologies enables secure, scalable, and democratized access to AI.
Architecture
Decentralized Marketplace on NEAR Blockchain2
This project builds a decentralized marketplace powered by NEAR Protocol and ValidCoin. The platform allows sellers to list AI/ML models for sale and buyers to purchase them using ValidCoin. Each transaction creates an escrow to ensure security, with funds released automatically after 7 days or upon reaching 1000 downloads. The system incorporates both marketplace and escrow functionalities, ensuring transparent and secure transactions while minimizing gas fees.
Log in or sign up for Devpost to join the conversation.