AI Tips

AI Tools for Software Development

This collection of tips provides practical guidance on using AI tools for students learning to code, the tips will be dispersed through the foundations material. These techniques complement the principles outlined in our Large Language Models and Integrity guide.

Research and Discovery

Learning New Technologies πŸ“ Used in: Sprint 1 - Version Control with Git, Sprint 1 - Command Line Primer

πŸ› οΈ Ask an LLM: "Explain [technology/framework] in simple terms with practical examples. What are the key concepts I should understand first?"

🎯 Why this works: Gets you oriented without overwhelming detail, focuses on fundamentals first.

⚠️Beware of: Over-reliance on explanations without being able to apply concepts independently.

Understanding Documentation πŸ“ Used in: Sprint 1 - Command Line Primer

πŸ› οΈ Ask an LLM: "I'm reading the documentation for [tool/library]. Can you break down this section into simpler language: [paste documentation]"

🎯 Why this works: Helps decode dense technical documentation into learnable chunks.

⚠️Beware of: Avoiding original documentation entirely, which prevents learning to read technical specs independently.

Technology Comparisons πŸ“ Used in: Sprint 2 - Technical Blog

πŸ› οΈ Ask an LLM: "Compare [Tool A] vs [Tool B] for [specific use case]. What are the trade-offs and when would I choose each?"

🎯 Why this works: Helps you understand not just what tools do, but when to use them.

⚠️Beware of: Accepting AI trade-off analysis without validating against real project requirements.

Error Resolution

Making Sense of Error Messages πŸ“ Used in: Sprint 1 - Asking for Help

πŸ› οΈ Ask an LLM: "I got this error message: [paste error]. Explain what this means and suggest 3 possible solutions, ranked by likelihood."

🎯 Why this works: Transforms cryptic errors into actionable steps and teaches pattern recognition.

⚠️Beware of: Trying solutions without understanding why the error occurred.

Getting Debugging Help πŸ“ Used in: Sprint 1 - Setup New Repo & Create Blog

πŸ› οΈ Ask an LLM: "I expect this code to [describe expected behavior] but it's [describe actual behavior]. Help me debug by asking clarifying questions about my code."

🎯 Why this works: Forces you to articulate the problem clearly, often revealing the issue.

⚠️Beware of: Over-relying on AI for hypothesis generation instead of developing your own debugging intuition.

Reading Error Stack Traces πŸ“ Used in: Sprint 3 - Manipulating the DOM

πŸ› οΈ Ask an LLM: "Here's my stack trace: [paste trace]. Walk me through reading this step by step and identifying the likely problem area."

🎯 Why this works: Teaches you to read stack traces independently for future debugging.

⚠️Beware of: Declining ability to read stack traces independently over time.

Code Understanding

Learning Different Approaches πŸ“ Used in: Sprint 3 - Intro to JavaScript

πŸ› οΈ Ask an LLM: "Show me 3 different ways to [accomplish task] in [language]. Explain the pros and cons of each approach."

🎯 Why this works: Exposes you to multiple patterns and builds understanding of trade-offs.

⚠️Beware of: Relying on AI patterns instead of learning idiomatic language constructs.

Getting Code Feedback πŸ“ Used in: Sprint 3 - JavaScript Cafe Challenge

πŸ› οΈ Ask an LLM: "Review this code and suggest improvements for readability, performance, and best practices: [paste code]"

🎯 Why this works: Gets feedback on your code style and introduces best practices.

⚠️Beware of: Treating AI suggestions as authoritative without understanding the underlying principles.

Recognizing Code Patterns πŸ“ Used in: Sprint 4 - Gradebook

πŸ› οΈ Ask an LLM: "What design pattern is being used in this code? Explain how it works: [paste code]"

🎯 Why this works: Helps you recognize common patterns and understand their purpose.

⚠️Beware of: Over-confidence in AI pattern identification without verifying against documentation.

Learning and Skill Development

Understanding Difficult Concepts πŸ“ Used in: Sprint 2 - Learn HTML & CSS

πŸ› οΈ Ask an LLM: "I'm confused about [concept]. Can you explain it using an analogy, then show a simple code example?"

🎯 Why this works: Multiple learning modalities help concepts stick.

⚠️Beware of: Shallow conceptual understanding that breaks down when you need to apply concepts in novel situations.

Creating Practice Problems πŸ“ Used in: Sprint 4 - JavaScript Kata

πŸ› οΈ Ask an LLM: "Create 3 coding challenges that would help me practice [specific skill]. Start with basic and increase difficulty."

🎯 Why this works: Provides targeted practice opportunities for skill building.

⚠️Beware of: Missing real-world complexity that AI practice problems often oversimplify.

Breaking Down Code πŸ“ Used in: Sprint 2 - JavaScript Primer

πŸ› οΈ Ask an LLM: "Explain this code line by line, including what each part does and why it's written this way: [paste code]"

🎯 Why this works: Builds deep understanding of how code works, not just what it does.

⚠️Beware of: Accepting AI explanations without verifying against official documentation or testing the code.

Best Practices for AI Assistance

Do:

  • Ask for explanations, not just solutions
  • Request multiple approaches to understand trade-offs
  • Use AI to clarify concepts you're struggling with
  • Ask follow-up questions when you don't understand
  • Verify AI suggestions with official documentation

Don't:

  • Submit AI-generated code without understanding it
  • Use AI as a replacement for learning fundamentals
  • Trust AI answers without verification
  • Ask AI to complete entire assignments for you
  • Rely on AI for factual information without double-checking

Power Prompts for Learning

The Teaching Prompt - AI as your pairing partner/tutor πŸ“ Used in: Sprint 5 - Pair Programming

πŸ› οΈ Ask an LLM: "Don't give me the answer. Instead, ask me questions that guide me to figure it out myself."

🎯 Why this works: Develops problem-solving skills rather than just providing solutions.

⚠️Beware of: Giving up too quickly when the AI asks challenging questions.

The Understanding Check πŸ“ Used in: Sprint 4 - Testing

πŸ› οΈ Ask an LLM: "I think I understand this concept. Quiz me with 3 questions to test my knowledge."

🎯 Why this works: Reveals gaps in understanding before you move on to harder topics.

⚠️Beware of: Only asking for easy questions that make you feel confident.

The Step-by-Step Helper πŸ“ Used in: Sprint 4 - Problem Solving and Pseudocode

πŸ› οΈ Ask an LLM: "Break down this complex problem into smaller steps I can tackle one at a time."

🎯 Why this works: Makes overwhelming problems manageable and teaches problem decomposition.

⚠️Beware of: Becoming dependent on AI to break down every problem for you.

The Context Builder πŸ“ Used in: Sprint 1 - Thinking Like a Programmer

πŸ› οΈ Ask an LLM: "Before we solve this, help me understand the bigger picture of how this fits into [larger concept/framework]."

🎯 Why this works: Builds connections between concepts and improves retention.

⚠️Beware of: Getting lost in too much context when you need to focus on specific skills.

The Alternative Explorer πŸ“ Used in: Sprint 5 - Tic-Tac-Toe

πŸ› οΈ Ask an LLM: "Show me 3 different ways to approach this problem, with the pros and cons of each."

🎯 Why this works: Develops critical thinking about different solution approaches.

⚠️Beware of: Analysis paralysis from too many options without trying any of them.

Remember

AI tools are most effective when used as learning partners, not solution providers. The goal is to enhance your understanding and skills, not to bypass the learning process. Always ensure you can explain and modify any code you use, regardless of its source.