Looking at the piles of AI Agent projects that have almost hit rock bottom (19B——>6B), there is a sense of collective death pressure. In addition to heartache, many people are confused. What happened to the AI Agent sector? After reflection and review, I would like to share a few opinions:
1) It is common to see piles of corpses in the early stages of a new trend, because the speed at which the troublemakers issue new assets is much faster than the speed at which innovative projects are implemented. Most people are forced to "PVP" in a sea of unproven concepts. Although we investment research bloggers try our best to discover valuable projects with a PVE mentality, we cannot escape the essence. The AI Agent narrative revolution is still a masked ball in the name of AI MEME.
However, the Fomo or pessimistic sentiment of retail investors cannot determine the final outcome. Once AI Agent develops a new Lego paradigm of sustainability, the choice of major funds will prove whether AI Agent can become popular.
2) AI Agent cannot be falsified, relying on the historical integration of AI + Crypto. In the short term, many people will criticize the concept of AI Agent in web3 based on the application value of AI Agent in web2, which is understandable. However, the core of AI Agent driven by applications will not change, except that web2 has chosen business models such as API calls and user payments, while web3 has chosen Tokenomics. Therefore, all abstract concepts above AI Agent will be verified in single AI application scenarios.
For example, for DeFai-like AI Agents, using TEEs to realize self-custody of assets is only the first step. Whether AI can autonomously convert natural semantic understanding into real operations such as Swap and Staking is the second step. However, the logic that truly verifies the maturity of the business model is the growth of TVL locked by AI Agents, the increase of TXs executed by users based on AI Agents, and the optimization of user UX experience (Gas friction, profit probability, etc.) by AI Agents executing on-chain transactions. Before these cannot be verified, it is difficult to be 100% sure which project will come out;
Before that, put aside the anxiety of missing out on MEME and lock in the benefits of innovative implementation of some valuable projects until a truly groundbreaking and verifiable business innovation point emerges and a mature "foundation" that can accumulate and enrich the commercial value of the portfolio is formed.
3) AI Agent seems to have been subdivided into many sub-fields including single AI, framework standards, DeFai, etc., but it has basically not gotten rid of the temperament of a "grassroots team".
Although any innovation field cannot escape the chaotic state of the "era of chaos" in the early stages, the lack of innovation in the products behind the #ai16z open source spirit totem, the challenges of mature agent landing behind the #Virtual innovation, and the market expectations of 30-300M that can be brought by a dev, a repo, and a potential business roadmap are obviously "irrational".
It should be said that the wrong valuation and pricing system, the wrong PVP speed pass Fomo environment, and the lack of confidence in value innovation have jointly caused the short-term collapse of AI Agent. Fortunately, the market expectations for AI Agent are still there. Although the iron-blooded holders are in pain, they are still holding on. Many entrepreneurial "regular forces" with strength, technical content and high barriers are rushing into the market, such as: #arc's continued pursuit and challenge of ai16z, etc.
Obviously, the early AI Agent web3 project model that relied on AI Agent narratives for rapid asset issuance and MEME-based gameplay has been falsified, and the next wave of innovation is gathering strength. This market cooling-off period may be the right time window for investing in R&D and exploring potential projects.
