vibe coding

"Vibe coding" is a modern, colloquial term for using natural language to write, debug, and iterate on code alongside an AI assistant. You no longer need to be a seasoned software engineer to build useful scripts—you just need to know how to logically explain what you want to achieve.

How AI Changes Workflows

AI drastically lowers the barrier to entry for coding tasks. Instead of remembering syntax or searching forums for boilerplate code, you act as the "architect" while the AI acts as the "typist," allowing you to build and automate much faster.

Rapid Scaffolding

Use Case: Starting a new script without staring at a blank page.

Example Prompt: "Write a Python script using pandas. It needs to pull JSON data from this public API endpoint, filter out inactive rows, and export it to a CSV file."

Instant Debugging

Use Case: Fixing broken code quickly.

Example Prompt: "I am getting a 'TypeError' on line 42 of this script. Here is the error traceback and the function code. Explain why it failed and rewrite it to fix the issue."

Supported Languages & Environments

Major AI tools (like Copilot, ChatGPT, and Claude) are trained on vast repositories of open-source code and excel at almost all common languages:

  • Web Development: HTML, CSS, JavaScript, React, Node.js.
  • Data & Scripting: Python, R, SQL, Pandas.
  • Automation: PowerShell, Bash scripts, Excel VBA, Google Apps Script.

Refactoring Code

Inherited a messy, uncommented script from a previous colleague? Paste it into the AI and ask it to:

  1. Explain what the code does step-by-step.
  2. Add helpful inline comments throughout.
  3. Refactor the logic to be more modern and efficient.

Language Translation

  • AI is excellent at translating logic between languages.
  • Example: Move a legacy statistics script from R to Python.
  • Example: Convert an old Excel VBA macro into a modern Office Script or Python function.

Chat Interfaces vs. Integrated Tools

While you can paste code into chat interfaces like Copilot or ChatGPT, advanced users may benefit from IDE integrations (like GitHub Copilot within VS Code). These tools sit directly in your programming environment, auto-completing lines of code as you type and reading your entire project directory for better context.

Security Reminder

Never include API keys, database passwords, or hard-coded credentials in your prompts. Ask the AI to set up environment variables for authentication. Additionally, always review and test AI-generated code in a safe environment before deploying it.

These pages are currently in beta. To give feedback, please contact DTS at ai.training@reading.ac.uk.