Inception Capital: Why did we invest in FLock?

Source: Inception Capital

Flock is an end-to-end AI co-creation platform that integrates decentralized federated learning with blockchain.

This cryptographically incentivized platform aims to democratize AI agent training, fine-tuning, and inference. It seeks to stop the collection of user data, protect privacy, and encourage broad governance participation. The result? Community-owned models built by many people, with data contributors and model builders also rewarded.

The AI ​​market is on a high growth trajectory, with a projected CAGR of 37.3% from 2023 to 2030, growing from $165 billion to $1.5 trillion. Revenues are already in the billions and are expected to reach nearly $60 billion in 2027.

But there are problems with the AI ​​value chain. The current market is dominated by highly centralized and closed-source companies, and geographic restrictions hinder billions of potential users and contributors. Companies like OpenAI are not open, and they currently receive the vast majority of revenue generated by AI applications.

To address this, Flock.io is building a co-creation hub for building new AI models and dapps customized by the community. Their platform empowers the community to build, train, and deploy AI models in a collaborative, privacy-preserving, and securely incentivized way, democratizing AI development.

By default, data is kept local, and model developers and data contributors are rewarded. Their goal is to decentralize the AI ​​value chain and maximize the power of the community.

They are also moving towards decentralizing the entire machine learning (ML) stack. From data to computing to training to inference, Flock has decentralized solutions in its roadmap.

Data:Train on local and crowdsourced data using federated learning, and use smart contracts to reward data contributions in a more fair and democratic way.

Compute:Run computations in any environment (decentralized computing).

Training:Out-of-the-box training clients enable Flock users to train ML models more easily and quickly. In addition, there is decentralized LLM fine-tuning and quality control with cryptoeconomic incentives and security protections.

Inference:Federated learning also enables decentralized model hosting, and smart contracts enable model developers to share revenue.

Flock has built a growing ecosystem of partners covering decentralized storage solutions (Filecoin/IPFS), healthcare (UCL Hospitals and Beijing Tsinghua Chang Gung Hospital), and Web3 infrastructure providers (Request Network). We believe they will continue to expand their ecosystem and integrations to achieve their vision of an AI future that is more democratic and open than today.

Flock's founder and CEO Jihao has extensive expertise in federated learning and a vision for how to combine blockchain and AI. He has demonstrated outstanding leadership and has successfully assembled an excellent team, including Oxford University alumni in computer science and mathematics, and an engineering leader with a PhD in computer science who previously founded Dola.ai (acquired by Tencent). We believe they will help build a future that optimizes the application of this protected data to develop proprietary AI solutions while protecting user privacy.

Welcome Flock!