Flock becomes the sole recipient of the grant from the non-profit foundation to further its academic research and build practical and secure blockchain-driven AI
LONDON, Sept. 3, 2024 /PRNewswire-PRWeb/ -- FLock.io, an industry leader in the decentralized AI (dAI) movement, became the sole AI infrastructure recipient of The Ethereum Foundation (EF) Academic Grants Round 2024. This grant will fund research into incentive mechanisms for smart contract-enabled blockchain solutions.
The Ethereum Foundation is a non-profit that supports the Ethereum ecosystem by funding protocol development, growing the ecosystem, and advocating for Ethereum. The Academic Grants Round 2024 is a $1M wave of grants to support Ethereum-related academic work.
FLock is at the forefront of federated machine learning on the blockchain. Having recently closed a $6 million seed round, it is building a world-first collaborative dAI training platform with on-chain governance, AI model training and hosting. The London-based Web3 startup, founded by a group of Oxford graduates, is on a mission to redirect power from centralized corporations into the hands of the community.
FLock.io is researching incentive mechanisms for blockchain-based AI training
This grant will accelerate FLock.io's pioneering research in optimizing incentive tokenomics mechanisms to deter malicious behavior in the decentralized AI development lifecycle. This project will benefit the greater ecosystem by harnessing smart contracts for secure dAI systems, upholding principles of security, trust, and computational integrity while maintaining data privacy.
The research, led by Dr. Zhipeng Wang, lead blockchain researcher at FLock, and Jiahao Sun, FLock Founder and CEO and Research Fellow at Imperial College London, will expand on the findings in FLockpublished paper, "Defending Against Poisoning Attacks in Federated Learning With Blockchain," in IEEE's Transactions on Artificial Intelligence journal. The research proposes a peer-to-peer voting and reward-and-slash system that protects against malicious activities by Federated Learning participants.
"We are immensely grateful for The Ethereum Foundation grants that support our research," said Dr. Wang, "We aim to investigate and establish a practical blueprint for leveraging the potential of smart-contract-enabled blockchains, particularly Ethereum, to support secure and decentralized AI systems."
"Our research seeks to extend the boundaries at the intersection of blockchain and AI ecosystems," Sun added. "We look forward to collaborating with diverse communities to revolutionize the landscape."
FLock is dedicated to supporting the decentralization movement. FLock is the first project to bring an end-to-end AI fine-tuning lifecycle to Web3, akin to a Web3 Hugging Face. Its permissionless, performant ML system comprises AI Arena, FL Alliance, and Model Marketplace. The AI Arena closed beta is underway, where developers compete on the best AI models on one validation dataset in a Kaggle-style leaderboard, using their own or decentralized compute.
Explore FLock.io's latest white paper, which dives into innovative approaches to decentralized machine learning, highlighting cutting-edge incentive mechanisms that leverage blockchain technology to enhance security, efficiency, and scalability in AI solutions. This comprehensive resource offers valuable insights for developers and industry professionals seeking to understand the future of AI-driven blockchain systems. Other existing academic papers co-authored by FLock include "zkFL: Zero-Knowledge Proof-based Gradient Aggregation for Federated Learning" and "GARNN: An Interpretable Graph Attentive Recurrent Neural Network for Predicting Blood Glucose Levels via Multivariate Time Series" in arXiv.
For future updates, follow FLock.io on Twitter, and see its blog and docs.
About FLock.io
FLock.io is a community-driven platform facilitating the creation of on-chain, decentralized AI models. By integrating federated learning and blockchain technology, FLock ensures equitable incentives for data contributors and fosters open collaboration. It also addresses the increasing need for advanced, bespoke AI models and reduces the risk of data breaches by providing secure model training without exposing source data.
Media Contact
Robert O'Neil, Flock.io, (310) 260-7901, [email protected], www.flock.io
SOURCE Flock.io
Share this article