Types of AI Coding Tools

GitHub Copilot and ChatGPT are both excellent AI tools, and it can be difficult to decide which one to use. However, it's important to understand that they are different types of AI. Here, we will focus on the actual differences in usage, rather than differences in governance or policy.

GitHub Copilot provides a feeling of working with a skilled engineer sitting next to you and coding together. On the other hand, a chat-based interface is like consulting with an expert one-on-one. The output may be slower and require leaving the editor, but with GitHub Copilot X's Technical Preview feature, GitHub Copilot for Chat, it can be used within the editor. However, it requires switching to the chat-style window.

Coding Together Style: GitHub Copilot

GitHub Copilot plays the role of a pair programming partner, suggesting appropriate actions during coding. Essentially, it quickly responds to rapid feedback. GitHub Copilot uses OpenAI's Codex model, which is a large-scale language model based on GPT3. It is lighter weight than GPT-3 and can provide faster responses.

Use cases for GitHub Copilot:

  • When rapid code completion or suggestions are needed for an existing codebase.

  • When there is already a clear idea for adding new features or components.

  • For simple coding tasks where results are desired quickly.

  • When there is existing similar code that can be used as a reference to generate new code.

Chat Style: ChatGPT/Bing/GitHub Copilot for Chat

Next, we will explain the chat style tools like GitHub Copilot X, Bing, and ChatGPT. These can be used when creating code or templates. At the initial stage, it may not be clear what needs to be created. In such cases, you can consult with AI like ChatGPT to complete the program while asking what kind of program to write in natural language.

There are pros and cons to each, but it's important to consider which one to use in different programming situations. Below are examples of use cases for each tool.

Use cases for Chat Style (ChatGPT/Bing/GitHub Copilot for Chat):

  • When consultation is necessary on specific implementation of code.

  • When it is necessary to consider ideas or direction in the initial stages of a project.

  • When more advanced explanation or questioning is necessary.

  • When step-by-step guidance is necessary to solve complex problems.

Tool selection

Ultimately, the choice of which tool to use depends on the needs and progress of your project. GitHub Copilot helps speed up coding, while chat-style AI is more suitable for deeper discussions or guidance.

Also, since these tools are not mutually exclusive, both can be utilized in different stages or situations of the project. For example, in the initial stage of the project, you can use chat-style AI to consider ideas and determine the direction, then use GitHub Copilot to efficiently code. This allows you to take advantage of the strengths of both tools to achieve optimal results.

Detailed differences

Below are the differences between each.

Item
GitHub Copilot (excluding GitHub Copilot X)
Chat AI (ChatGPT/Bing)

Use cases

Mainly for code writing and small-to-medium code completion

Code generation, template creation, and can provide context for generating relatively large code through natural language

Interface

Editor

Browser

Response speed

Fast

Somewhat slower

Natural language understanding

High

Higher

Code snippet generation

Powerful. Tends to work better for shorter code generation than longer code generation.

Supports both long and short code

Programming language support range

Wide range

Wide range

Real-time auto-completion and suggestion

Real-time

Limited

Text generation

Possible in high dimensions, but inferior to GPT3

Powerful

Code review

Partial support

Suggestions and comments through natural language

Debug support

Limited support, but requires ingenuity

Can propose debugging steps and approaches

Code style and convention application

High

Requires passing more context beforehand

Checklist

Last updated