Clickable Options in the Editor - The fundamental idea is to present options that users can click on, rather than having the AI make decisions automatically. This approach gives you control over the development process and helps you understand the trade-offs before committing to code. Download the thesis.
Traditional AI coding assistants often generate code immediately, which can lead to:
The goal is to create a coding assistant that feels like a collaborative partner rather than an autocomplete tool. By presenting options and explaining reasoning, we aim to:
This demo showcases the core concept, but the approach could be extended with:
Instead of immediately generating code, OptionPilot presents you with decision points that break down complex problems into manageable choices:
When you ask a coding question, the AI identifies key decision points and presents multiple implementation options. Each option includes:
After you've navigated through the options and made your choices, you explicitly request code generation. This ensures:
Every code snippet is accompanied by explanations that articulate:
Complex problems are broken down into clear decision points with multiple options, preventing information overload.
You can explore different options, see their implications, and refine your choices before generating code.
Each option includes pros and cons, helping you quickly understand trade-offs between different approaches.
The AI works with your code context (like the TaskManager in this demo) to provide relevant, applicable suggestions.
Common decision paths are preloaded in the background, making interactions feel instant when you click on options.
You decide when to generate code, which options to explore, and how deep to go into each decision point.
By presenting decisions incrementally rather than all at once, we help you process information more effectively.
You're in charge of the development process. The AI guides and suggests, but you make the final decisions.
Every suggestion includes explanations that help you understand not just what the code does, but why it's structured that way.
By exploring options and understanding trade-offs, you build mental models that improve your coding skills over time.