The Potential of AI Data Tracks and the Rise of Web3 DataFi

In an era where the world is competing to build the best foundational models, while computing power and model architecture are crucial, the real competitive advantage lies in the training data. The most notable event in the AI circle this month is Zuckerberg forming a Meta AI team primarily composed of Chinese research talent, led by 28-year-old Alexander Wang. Wang's company, Scale AI, is currently valued at $29 billion and provides data services for several AI giants.

The reason Scale AI stands out among numerous unicorns is that it early on recognized the importance of data in the AI industry. Computing power, models, and data constitute the three pillars of AI models. As most models adopt the transformer framework, the importance of data has become increasingly prominent after major companies solved the computing power issue.

Model training is divided into two stages: pre-training and fine-tuning. The pre-training stage requires a large amount of text, code, and other information crawled from the web, while the fine-tuning stage requires carefully processed and targeted datasets. These two types of datasets constitute the main body of the AI Data track. As the model's capabilities improve, high-quality, professional training data will become a key factor in determining the model's performance.

Compared to traditional data companies, Web3 has a natural advantage in the AI data field, giving rise to the concept of DataFi. The advantages of Web3 DataFi include:

1. Smart contracts ensure data sovereignty, security, and privacy.
2. Distributed architecture attracts the most suitable workforce globally
3. Blockchain provides clear incentive and settlement mechanisms.
4. Conducive to building an efficient and open data market

For ordinary users, DataFi is the best entry point for participating in decentralized AI projects. Users can get involved by providing data, evaluating models, and other simple tasks, making it relatively accessible.

Recently, several Web3 DataFi projects have secured considerable funding. The main projects include:

• Sahara AI: Building a decentralized AI infrastructure and trading market
• Yupp: AI Model Feedback Platform
• Vana: Transform personal data into monetizable digital assets
• Chainbase: Focus on on-chain data
• Sapien: Transforming human knowledge into AI training data
• Prisma X: Open Coordination Layer for Robots
• Masa: The data subnet project of the Bittensor ecosystem
• Irys: Programmable Data Storage and Computing
• ORO: Empowering ordinary people to participate in AI contributions
• Gata: Decentralized Data Layer

Although these projects currently have low barriers to entry, early accumulation of users and ecological stickiness is crucial. The challenges they face include how to manage manpower, ensure data quality, and improve the transparency and decentralization of the projects.

The large-scale application of DataFi requires attracting enough individual users to participate in the data ecosystem, while also gaining recognition from mainstream enterprises. Some projects like Sahara AI and Vana have made good progress in this regard.

DataFi represents the long-term symbiotic relationship between human intelligence and machine intelligence. For those who feel anxious in the AI era or still hold onto blockchain ideals, participating in DataFi may be a wise choice that aligns with the trend.