Mission

TuringNet is a public chain that provides open and trustable platform for AI.

The platform sets future development and applications of Artificial Intelligence free from any controls of centralized organizations, thus serving the best interests of the AI community that reshapes humanity.

Problem

  • Centralized model ownership

    Today’s existing AI models are increasingly owned and controlled by tech giants who offer cloud-based products and APIs. Such ownership could result in monoplication, unfair service prices and limit participation of smaller companies to innovate the industry in the long term.

  • Wasted computational resources

    Typical AI system requires massive computation powers for model training and inference. Multiple centralized systems have built their own siloed data centers, which could lead to enormous resource waste. There is a great opportunity to organize computational resources from public in a decentralized but trustable way.

  • Lack of incentives to train AI models

    Smaller companies and ordinary users are not incentivized participate in advancing AI industry as there is no fair share of rewards and a high bar to entering such market. Tech giants are not obligated to pay contributors who want to contribute their computation powers, as they possess large data centers with much lower cost per unit power. The ecosystem then separates the public from participating, and further strengthens centralized AI systems and models.

  • Expensive operational cost

    Centralized AI systems are operated separately by their owners. Even though the cost per unit power is lower, the scales of the systems often lead to enormous cost to run.

  • Data storage and exfiltration

    Training and optimizing AI models require large amount of data. The primary sources of such training data are from user characteristics and interactions, or copied from enterprise clients by providing cloud-based solutions. But there has been an increasing risks and concerns of data sharing and exfiltration without authorization.

Solution

TuringNet is creating world’s first blockchain network with consensus, incentives, extensibility, and virtual machine features dedicated for an open and trusted AI platform.

solution
LBFT Consensus Tailored for AI

The LBFT(Learning Byzantine Fault Tolerance) consensus mechanism is the hallmark of TuringNet’s innovation. It was specifically designed for AI model training and empowers seamless integration between AI tasks on blockchain within every iteration of consensus. By voting the minimum loss value in each iteration, the consensus is able to select the best performing node and confirm it with token rewards.

solution
Building Trust with Next Generation DModel and Multi-layer Chain

TuringNet’s decentralized models (DModels) are trained in an open environment and are 100% accessible and verifiable by the public. Its next generation multi-layered architecture is optimized for fast speed and large scale data sets for a wide variety of AI tasks, both in training and inference.

TuringNet Token (TNET)

TNET Token is a sole crypto currency in TuringNet’s ecosystem, and a functional utility token in value exchange and transition network. The TNET Token will not be issued by TuringNet, but 100% distributed to miners who accomplish mining tasks on TuringNet’s platform operated by smart contract.

Documents

Learn more details
FAQ

Community

Join our community today to receive news and updates