✍️ If you are optimistic about AI agents, you should pay attention to AI oracle machines The oracle has always been a track that is not sexy but has great opportunities, and the most typical case is LINK. In 2019, I bought a 1-U LINK, but at that time, I lacked knowledge of the oracle and felt that it was difficult for me to pay for this boring narrative. I made a small profit and ran away. Later, the oracle race broke out, and the leading LINK continued to soar, reaching a peak of over 50 U. At that time, although I couldn't understand it, I was greatly shocked. After many years, the oracle racing track is still a hot spot for entrepreneurship, but none of the rising stars can challenge LINK. Most projects have a low presence for individual investors, and their participation is also very limited. Recently, @ DeAgentAI launched the zkTLS AI oracle, which generates ZK proofs to verify the authenticity and privacy of API calls, helping the rise of AI Agent economy and potentially bringing opportunities for secondary participation. From a conceptual perspective, the AI oracle may seem a bit obscure and difficult to understand, but in practical applications, it has great potential and even has the potential to open up a new track, similar to Link in the past. Simply put, traditional oracle machines are designed for simple, public, and easily verifiable data, and have performed exceptionally well in the "public data" market. However, when the data becomes complex, private, and non deterministic, this model is prone to collapse. In the field of AI agents, more and more agents are emerging, but privacy and trust are still challenges. On the one hand, agents lack trust in off chain data and are unable to make advanced decisions based on richer data; On the other hand, different agents previously lacked trust and were unable to perform advanced operations based on trust. The above problems are difficult to solve through traditional oracle machines, as they are essentially "reputation consensus", meaning node voting and data disclosure. However, such solutions cannot meet the privacy needs of AI, nor can they make decisions based on complex data. DeAgentAI is targeting this pain point. However, they did not build a "more trustworthy" oracle, but rather a oracle that does not require trust at all: by creating a "cryptographic notary" system through zero knowledge proofs. The system provides identity verification, data integrity, and decision traceability for agent interactions through "off chain proof - privacy protection execution - proof generation - on chain verification" while protecting privacy. If the recent popularity of x402 has solved the friction of AI payments, DeAgentAI has provided "verifiable trust", making it possible to create a truly autonomous, trustworthy, and high-value AI agent economy. The emergence of AI oracle machines is not to compete with traditional oracle machines such as LINK, but to open up a new market. It relies on the huge AI agent economy to solve new problems in the AI era, and its value will become increasingly prominent. For DeAgentAI, they are already the largest AI agent infrastructure across Sui, BSC, and BTC ecosystems (18.5 million users, peak 440000 daily active users, 195 million on chain transactions), and look forward to dominating the AI oracle market in the future. I still hold some AIA and hope to take off.