LAB | The AI Assistant Trials, who’s the Best Co-Pilot?

🧪 Focus: Critical AI comparison, Prompt Engineering, Output Review

🧠 Scenario

Your dev team is considering integrating an AI assistant into your daily workflow, but which one fits best?

You’ve been tasked with leading “The AI Assistant Trials”, a side-by-side comparison between:

• ChatGPT (Free)
• Claude (Free)
• Your pick (if you use another AI tool like Gemini or so)

You’ll evaluate each AI by giving them the same set of coding challenges and analysing their responses.

🎯 Learning Goals

By completing this lab, you will:

✅ Compare AI assistants on real-world tasks
✅ Practice prompt clarity and iteration
✅ Learn how to critically evaluate AI suggestions
✅ Choose the best assistant for your own dev flow

Requirements

• Fork this repo.
• Clone this repo.

Submission

• Upon completion, run the following commands:

git add .
git commit -m "Solved lab"
git push origin master

• Create a Pull Request so that your work can be checked.

🧪 Tasks

Perform 3 challenges with each AI. You must use the same prompt for all three.

✅ Task 1: Explain a Function

Prompt:

“Explain this JavaScript function in detail. Add comments as needed.”

function mysteryOp(arr) {
  return arr.reduce((acc, val) => acc ^ val, 0);
}

✅ Task 2: DOM Manipulation

Prompt:

“Write a function that changes the text of an HTML element with id status when a button is clicked. Use vanilla JavaScript and follow accessibility best practices.”

✅ Task 3: Refactor Obfuscated Code

Prompt:

“Refactor this function to make it readable and efficient. Add meaningful variable names and comments.”

function r(a){let b=0;for(let i=0;i<a.length;i++){if(a[i]%2===0){b+=a[i]}}return b}

📊 AI Evaluation Table

AI Tool	Task	Clarity (1–5)	Accuracy (1–5)	Speed (1–5)	Tone (1–5)	Notes
ChatGPT	1
ChatGPT	2
ChatGPT	3
Claude	1
Claude	2
Claude	3
Your Pick	1
Your Pick	2
Your Pick	3

💡 Don’t just give scores write down what stood out.
Was one clearer, but verbose? Another too short but correct?

🧾 Your Recommendation Report

Create a Markdown report in your project folder with the following sections:

AI_Assistant_Report.md

AI Assistant Trials – Final Report

🏆 My Pick:

[Name of AI]

✅ Pros & Cons

ChatGPT

• ✅ [Insert good things]
• ❌ [Insert limitations]

Claude

• ✅
• ❌

[Your Pick]

• ✅
• ❌

📌 Task-by-Task Highlights

• Task 1: [Summary of performance]
• Task 2: ...
• Task 3: ...

🔍 Surprises & Bugs

• [Mention hallucinations, bugs, or bad suggestions]

Final Thoughts

Which AI would you trust in a real project? Why?

Reflection Prompts

Write in your journal or digital notebook:

• Which AI helped you learn better?
• Was there a big difference in explanation quality?
• Which one matched your communication style?
• How might you use different AIs for different types of work?

🔬 Bonus Round

Ask all three:

“What are 3 common mistakes junior developers make in DOM manipulation, and how can they be avoided?”

Who gave the clearest, most useful guidance?

✅ Submission Checklist

• Filled-out evaluation table
• AI_Assistant_Report.md with a clear winner
• Reflections written

🎯 You’re the AI Captain Now

Well done, dev! You’ve tested, documented, and decided.

Remember: These tools are powerful but you are the architect of your process.

Choose wisely. 👨‍💻 👩‍💻