What is Vibe Coding?

A programming approach where developers describe what they want in natural language and let AI generate the code.

🎵

Definition

Vibe Coding is a programming approach coined by Andrej Karpathy in 2025, where developers describe what they want in natural language and let AI models generate the actual code. The developer focuses on the "vibe" -- the intent and direction -- while AI handles implementation details.

🎯

How It Works

  • The developer describes the desired functionality in plain language
  • An AI coding assistant (like Claude, Cursor, or Copilot) generates the code
  • The developer reviews, tests, and iterates using conversational feedback
  • The focus shifts from writing code to directing and validating AI output
💡

Practical Example

Instead of manually writing a REST API with authentication, a developer tells the AI: "Create an Express API with JWT authentication, user registration, and login endpoints with input validation." The AI generates the complete implementation, which the developer then reviews and refines.

⚖️

Tradeoffs

  • Pros: Faster prototyping, lower barrier to entry, focus on problem-solving over syntax
  • Cons: Risk of accepting code you don't understand, potential security issues, dependency on AI quality
📈

Why It Matters

Vibe Coding represents a fundamental shift in how software is built, making programming more accessible while requiring developers to develop new skills in code review, prompt engineering, and AI collaboration.

🍄

Want to learn more?

If you're curious to learn more about Vibe Coding, reach out to me on X. I love sharing ideas, answering questions, and discussing curiosities about these topics, so don't hesitate to stop by. See you around!