A Comparative Analysis of the Development Paths of AI and Crypto Assets Technologies

Recently, Ethereum's Rollup-Centric strategy seems to have encountered setbacks, and the development model of multi-layer architecture (L1-L2-L3) has also come under scrutiny. However, interestingly, the field of artificial intelligence has also experienced a similar rapid evolution across multiple levels in the past year. By comparing the developmental trajectories of these two fields, we can delve into their similarities and differences.

In the field of AI, each layer of the multi-layer architecture is dedicated to solving the core problems that the previous layer failed to tackle:

1. Large language models (LLMs) on the L1 layer lay the foundation for language understanding and generation, but there are obvious shortcomings in logical reasoning and mathematical calculations.

2. The reasoning model of the L2 layer specifically addresses these issues. For example, some advanced models have already been able to handle complex mathematical problems and code debugging, effectively compensating for the cognitive blind spots of LLMs.

3. The AI agents at Layer 3 integrate the capabilities of the first two layers, transforming AI from passive response to active execution, enabling it to autonomously plan tasks, invoke tools, and handle complex workflows.

This layered structure reflects the progression of capabilities: L1 lays the foundation, L2 addresses shortcomings, and L3 achieves integration. Each layer realizes a qualitative leap based on the previous layer, allowing users to clearly feel that AI is becoming more intelligent and practical.

In contrast, the multi-layer architecture in the Crypto Assets field seems to face different challenges:

1. The performance limitations of L1 public chains have led to the emergence of L2 scaling solutions. However, despite reduced gas fees and improved TPS, new problems such as liquidity fragmentation and a lack of ecological applications have also arisen.

2. The emergence of L3 vertical application chains aims to address the issues of L2, but has led to further fragmentation of the ecosystem, making it difficult to enjoy the synergies brought by common infrastructure.

This layering seems to have become a "problem transfer": the bottleneck of L1 leads to the emergence of L2, the issues of L2 in turn give rise to L3, where each layer merely transfers the problem from one domain to another, rather than truly addressing the fundamental issues.

The core reason for this difference may lie in the following: the stratification in the AI field is driven by technological competition, with major companies striving to enhance model capabilities; whereas the stratification in the Crypto Assets field seems to be more dominated by token economics, with core metrics for each layer often focused on Total Value Locked (TVL) and token prices.

This comparison reveals the starkly different development drivers in the two fields: one focuses on solving technical challenges, while the other emphasizes designing financial products. Although this abstract comparison is not absolute, it indeed provides us with an interesting perspective, allowing us to think about these two rapidly evolving technological fields from different angles.