📄 中文摘要
ThinkLang 是一种创新的开源编程语言，旨在将 AI 模型调用集成到语言核心，使其像声明变量一样自然。该语言通过内置的 `think`、`infer` 和 `reason` 等关键字，消除了传统 AI 应用开发中常见的库导入、API 封装和 SDK 调用等繁琐步骤。ThinkLang 将代码转译为 TypeScript 并在运行时调用大型语言模型（LLM），但其编写体验与直接调用 API 截然不同。这种设计理念解决了现有 AI 应用开发模式中，开发者需要反复编写提示、调用 API 并解析响应的痛点。通过将 AI 能力提升为语言原语，ThinkLang 旨在简化 AI 应用的开发流程，提高开发效率，并使 AI 功能的集成更加无缝和直观。该项目已开源，展示了其构建初衷、具体实现和未来发展方向，预示着一种全新的 AI 编程范式。
📄 English Summary
Introducing ThinkLang: A Programming Language Where AI Is a First-Class Citizen
ThinkLang is an innovative open-source programming language designed to integrate AI model calls as first-class citizens, making them as natural as declaring a variable. This language eliminates the typical boilerplate of library imports, API wrappers, and SDK calls by incorporating keywords like `think`, `infer`, and `reason` directly into its core primitives. While ThinkLang transpiles to TypeScript and invokes Large Language Models (LLMs) at runtime, the development experience fundamentally differs from wiring up API calls. This approach addresses a common pain point in current AI application development, where developers repeatedly write prompts, call APIs, and parse responses. By elevating AI capabilities to language primitives, ThinkLang aims to streamline the development process, enhance efficiency, and foster more seamless and intuitive integration of AI functionalities. The project has been open-sourced, detailing its motivation, implementation, and future trajectory, heralding a new paradigm in AI programming.