MuleRun CTO Shu Junliang: Model gap quickly converges, Agent moat turns to 'speed+data' BlockBeats News: On April 21st, during a roundtable discussion on "Cracking Web 4.0: When AI Agents Take Over Chain Permissions," MuleRun CTO Shu Junliang discussed the topic of "Agent moats," stating that the traditional AI technology moat is rapidly weakening due to the accelerated convergence of model capability gaps and exponential improvement in development efficiency. He pointed out that the performance gap between mainstream models is rapidly narrowing, especially in the past year when the gap in model capabilities between domestic and foreign models has significantly converged. Meanwhile, with the explosive improvement of coding capabilities, software development efficiency has significantly increased - functions that used to take weeks or even months to complete can now be achieved within days. This means that both the Agent framework and specific functional modules can be quickly reused or replicated through open source solutions, making the product level 'functional moat' increasingly fragile. In this context, Shu Junliang believes that the core competitiveness of future agents will mainly be reflected in two aspects: one is the ability to continuously iterate at high frequencies, that is, whether the team can maintain a leading product update speed in the long run; The second advantage is data, including exclusive data resources and user generated data. On the one hand, platforms with unique data acquisition capabilities (such as industry-specific or regional data) will form natural barriers; On the other hand, as users accumulate behavior and memory data through long-term use on the platform, it will also become a key asset that is difficult to migrate, further enhancing user stickiness and product competitiveness. He concluded that under the trend of "equalization" of models and technologies, the moat of agents is shifting from "technical capabilities" to a comprehensive competition of "data assets and execution efficiency".