Skip to content

AI-DISCLOSURE.md

Latest commit

 

History

History
81 lines (64 loc) · 4.06 KB

AI-DISCLOSURE.md

File metadata and controls

81 lines (64 loc) · 4.06 KB

AI Usage Disclosure

Content

Why this document exists

Transparency about AI tool usage matters. This document:

  1. Describes and clarifies how and when I use AI in this project
  2. Serves as a guidance for contributors

My use of AI

Rationale

The primary reason I'm doing all of this is to learn, understand, and grow my skills in programming, debugging, and troubleshooting. If I'm not learning something new or interesting from a task, that's when I use AI.

I maintain full control over creative decisions. Everything else that's far outside my skillset is fair game.

Data privacy: I never provide sensitive data to AI tools. Everything is anonymized.

Attribution: I don't explicitly mark AI-generated code because I always modify and format it anyway. The final code is mine.

Examples

Design: As a full-time designer, I have full creative control over this project. It's the thing that brings me joy and I would never outsource this to any AI.
Problem Solving: Problem-solving, debugging and troubleshooting is where the learning happens, as it helps me to understand how things work more clearly, so I do that myself.
Documentation: I know how to write, but I don't need to craft every sentence in project documentation from scratch. I use AI to get a springboard/boilerplate, which I then heavily modify to fit my voice and my needs.
Scripting: Since this area is outside my skillset (for now), I rely on AI tools to help me.

Tools

  • Claude AI - Brainstorming ("what's the best approach to X") and code bugfixes
  • ChatGPT 5 - One-off feedback on minor code details or phrasing

What AI does

  • Scripts - Help with implementing functionality outside my skillset
  • Debugging - Last resort after exhausting documentation and other sources
  • Troubleshooting - Discussing approaches to solving problems (primarily Claude AI)
  • Documentation feedback - Suggestions on phrasing and style (I write 90% of the content myself)

What AI doesn't do

  • Design work - Any design or creative decisions
  • Commits - AI doesn't write my commit messages
  • Code review - AI doesn't review code
  • Project ideation - Coming up with project concepts or features

Workflow

When I encounter a problem:

  1. Try to solve it myself first
  2. Peruse project documentation
  3. Search alternative sources (Reddit, StackOverflow)
  4. If everything fails, I use AI in this manner:
    • I define the problem, provide context, explain what I want and don't want, set expectations
    • AI provides a springboard, initial draft or solution based on my requirements
    • I take over, modify, improve, or completely rewrite the draft; I ask for feedback if needed
    • AI gives feedback, suggestions or validation
    • I make the final call, process the feedback (or don't) and finalize the result

Key point: I'm driving the process. AI accelerates specific steps, but doesn't replace my thinking or decision-making.

Responsibility

  • AI is a tool, I'm the author - I'm responsible for reviewing and maintaining any AI-assisted code, with the goal of eventually replacing it with human-written code
  • Users are responsible - Users are responsible for how they use this setup

For contributors

Please don't reach for AI as your first solution. AI-generated code without proper understanding creates technical debt and makes the codebase harder to review and maintain.

Recommended approach:

  1. Work on the problem yourself first
  2. Research and understand the solution space
  3. Use AI as support, not as a replacement for thinking
  4. Always review and understand what AI suggests before implementing it

The goal is maintainable, understandable code -- not code that somehow works.