# A lot of work around AI in 2023 was spent on building picks and shovels. I would know this because t...

Canonical URL: https://www.traeai.com/articles/27cdf921-84e3-42dd-afea-4045efb68f18
Original source: https://x.com/jerryjliu0/status/2050373987860123971
Source name: Jerry Liu(@jerryjliu0)
Content type: tweet
Language: 中文
Score: 7.8
Reading time: 3 分钟
Published: 2026-05-02T00:37:07+00:00
Tags: AI, LLM, Agent, LlamaIndex, Document Parsing

## Summary

Jerry Liu指出2023年AI基建（如LlamaIndex）聚焦于底层工具建设，而当前重心已转向稳固的智能体抽象与编码能力；真正持久的护城河在于构建支持智能体行动的上下文层，尤其是复杂文档解析与结构化。

## Key Takeaways

- 2023年AI工作重心是打造‘铲子和镐’式基础设施，LlamaIndex早期即为此定位。
- 核心智能体抽象已趋稳定，编码智能体正推动更高阶的智能体工程范式。
- 智能体的长期价值壁垒在于上下文层建设——尤其文档解析、工具集成与系统对接能力。

## Outline

- 背景：2023年的AI基建浪潮 — 回顾2023年大量投入集中在底层工具链，LlamaIndex最初使命即为此类‘铲子和镐’建设。
  - 范式迁移：从基建到智能体抽象 — 当前核心智能体抽象已固化，编码智能体使开发者能以更高抽象层级构建智能体。
  - 持久护城河：上下文层 — 智能体需依赖上下文才能执行任务，其关键组件包括Web搜索、系统记录、SaaS工具等。
    - LlamaIndex的聚焦点：文档上下文基建 — 专注将复杂PDF等非结构化文档精准解析为高质量语义上下文，并扩展配套工具链。

## Highlights

- > A lot of work around AI in 2023 was spent on building picks and shovels. — 原文首句
- > The core agent abstractions have solidified, and coding agents are letting everyone 'engineer' agents at a higher level of abstraction. — 原文中段
- > What continues to have a durable moat is building the context layer for agents. — 原文关键判断句
- > We're building the best-in-class capabilities for agents to parse the most complex documents into clean context. — 原文末段

## Citation Guidance

When citing this item, prefer the canonical traeai article URL for the AI-readable summary and include the original source URL when discussing the underlying source material.