Encryption AI Agent Development History and Future Outlook

In October 2024, with the emergence of the truth terminal ($GOAT), the wave of encryption AI agents officially began. In the following three months, this field experienced a frenzy of prosperity, with the market value once exceeding 10 billion dollars. Although there were very few substantial results left during this period, we did see various possibilities for the future development of encryption AI agents.

New experimental projects emerge every week, with some reaching a market value of 50 million dollars in a short period. The encryption AI framework and launchpad are touted as "AI agent L1." Some ChatGPT-style interface projects with a market capitalization of over 500 million dollars claim to be capable of executing on-chain operations, but in reality, they have no substance other than demonstrations. There are also projects that promise to create autonomous hedge funds, but they have nothing but white papers.

However, the period from February to April 2025 is a difficult time for encryption AI agents. The issuance of the TRUMP token has siphoned off most of the liquidity, resulting in a sharp decline in the market value of most tokens by 50-90%. The development progress of many teams has noticeably slowed, with over 90% of teams halting their work. Some top projects like zerebro suddenly suspended operations, griffain stopped releasing demos, while ai16z turned to chasing Meme narratives.

Despite the current grim situation, there is still some hope in the encryption AI field. In the next 6-12 months, we may see the following developments:

1. Optimization of intelligent chat applications. Although the current model still has shortcomings in tool invocation and understanding the context of encryption, the situation is improving. The latest large language models perform better in proactive operations and generative UI.

2. Consumer-level encryption AI applications based on vibe encoding. AI allows anyone to create their own applications or agents simply by inputting prompts. The combination of tokens and AI may give rise to interesting consumer platforms.

Looking towards a farther future, we may see:

1. AI agents become the "Trojan horse" for the popularization of stablecoins. AI agents with wallets may tend to use stablecoins for payments, which could promote stablecoins to become "first-class citizens" in the payment field.

2. AI technology is embedded in all encryption protocols. Most encryption protocols may begin to incorporate AI features more, making them smarter and more automated.

3. A coordinated network of encryption economy built for AI/agents. The advantages of cryptocurrency in coordinating capital and incentives may be used to build infrastructure that supports AI agents.

4. Composable personal context layer. The composability of blockchain may combine with AI's contextual understanding capabilities to create new user experiences and business models.

5. A chat-based encryption super application. As user experience shifts towards intention-based approaches, chat interfaces could become an ideal way to integrate various encryption services.

Despite the current market environment being filled with challenges, innovation in the encryption AI field has not stopped. We are at a critical stage in building the next generation of network forms, which is the best time for innovation and development.