Recently, many AI-related projects have emerged on the blockchain, such as DARK, MCP, and others. It feels like the market for AI-related projects is warming up again. Personally, when it comes to AI, I feel that besides the model itself, the dataset used for the model is also extremely important. For example, Grok's success is not just about its model but also because it has access to the entire Twitter dataset for training. From the perspective of datasets, I looked into whether there are any innovative projects that stand out, and I came across Irys @irys_xyz. Irys is the first Layer-1 blockchain designed specifically for AI that integrates data storage, execution, and monetization into a single high-performance network. It also enables data programmability. AI is accelerating the creation of data, and Irys is addressing the data challenges brought by AI. Today, the volume of data being generated and used is enormous. On the supply side, there is a lack of mechanisms for data ownership and monetization. People create data every day, and this data is used to train AI models, but the value generated from it only flows to a few large tech companies. On the demand side, the rise of AI has clearly shown that the future narrative of AI will heavily depend on the data that supports its operation. Ultimately, the network that wins the AI data battle will not be the one that merely "stores" data but the one that facilitates the generation of high-quality data and returns the value of that data to users. On-chain data is verifiable, instantly accessible, and programmable. Protocols can incentivize the management of high-quality datasets through automated payments. Irys' AWS model enables computation, databases, and applications to work together. For example, in a data service on Irys: Users upload data → The database protocol indexes it → The monetization protocol licenses it → AI models and applications use it in real-time. This essentially creates a flywheel effect: More data → More utility → More adoption → More monetization → Accelerated cycle. Through this model, data storage and smart contract execution are natively integrated, allowing projects to easily interoperate and share data. Additionally, because automated instructions are embedded in the data, tasks such as payments, licensing, and smart contract triggers can be executed. This enables instant, low-cost data retrieval with stable and predictable pricing. The potential market for data is at least $3 trillion. However, there is a significant amount of hard-to-access, low-structured data. Irys' AWS project can generate $110 billion in annual revenue and is valued at $1.5 trillion. Not to mention the growing demand for structured, high-quality data. By 2032, the entire market is estimated to reach $170 trillion, indicating a very promising future. Honestly, Irys addresses the core issues of current Web3 data management, such as Filecoin's slow retrieval, Arweave's high costs and instability, and Ethereum blob's insufficient storage capacity. Being the first Layer-1 blockchain explicitly designed for data storage and execution does sound quite appealing. Recently, Irys launched a new Chinese Twitter account @cn_irys_xyz. It's worth following.