FLock.io, a platform for on-chain decentralized AI model creation, is partnering with DePIN network IO.Net, to enhance its capabilities with decentralized compute power. The projects are joining forces to improve decentralized AI training platforms through advanced decentralized computing. This partnership aims to address the critical vulnerabilities of centralized AI systems by promoting a distributed approach in governance and computation.
The collaboration aligns with the broader industry trend towards decentralization, seen as essential for reducing risks such as data monopolization and privacy breaches. By integrating IO.Net’s decentralized compute resources, FLock aims to offer more robust AI solutions that are community-driven and less prone to the pitfalls of central control.
FLock utilizes federated learning for AI model training, where models learn from decentralized data sources without relocating the data, maintaining privacy and reducing the risk of misuse. IO.Net contributes by harnessing idle computing power, offering significant cost savings compared to traditional cloud providers.
Ahmad Shadid, CEO of IO.Net, shared his insights on the technological synergy between the two firms. “This partnership not only makes decentralized computing power more accessible but also marks a step forward in creating a resilient framework for AI development,” said Shadid.
FLock.io allows AI models to be trained directly on user devices, ensuring that sensitive data does not leave its source. This approach protects privacy and harnesses a diverse range of data, potentially leading to more accurate and representative AI models.
Jiahao Sun, CEO of FLock.io, further elaborated on the economic implications of decentralized AI. “While decentralized solutions are often viewed as cost-saving measures against traditional cloud services, their true value lies in their capacity to handle sensitive data with enhanced privacy,” he explained. Sun argued that the current efficiency of centralized platforms would eventually be outweighed by decentralized models’ benefits in privacy and specialized performance.
Sun asserts that decentralizing AI is as critical as the decentralization movement observed in finance. He believes that AI’s future lies in decentralized approaches that enhance privacy and model accuracy, particularly in sensitive sectors like finance and healthcare.
He told CryptoSlate,
“I believe that decentralized AI offers benefits in both cost reduction and safeguarding against the risks of centralized AI, especially frontier AI[…]
The true unrivaled benefit of decentralized AI lies in its potential to access private data in completely privacy-preserving ways. This enables AI to serve industries and sectors that were previously difficult for centralized third-party AI training providers to reach, such as finance and healthcare.”
Decentralized AI can access private data without compromising privacy, presenting a potential for superior model performance across diverse domains, according to Sun. “By integrating blockchain technology, we can facilitate a shift in AI model training that emphasizes community involvement and data security,” added Sun.
The conversation around decentralized AI is becoming increasingly relevant as the technology advances. Sun’s vision for FLock.io involves adapting AI to operate on decentralized infrastructure and rethinking how AI can be more inclusive and secure. “We are setting the stage for future applications where decentralized AI could lead to breakthroughs in industries previously hampered by data privacy concerns,” Sun concluded.
As the partnership progresses, FLock and IO.Net will continue to explore how decentralized computing can revolutionize AI, making it more adaptable, private, and aligned with user-centric governance. Sun predicted that a move toward locally run decentralized AI models
“has the power to reshape the AI landscape and pave the way for groundbreaking applications that were once considered impractical or unfeasible.”
The post FLock to train decentralized AI models with IO Net to reduce centralized privacy risk appeared first on CryptoSlate.
