The ChatGPT mistake that weakens your engineering skills

What this guide covers

You’ll learn how to use ChatGPT as a drafting tool while relying on your own expertise to verify, refine, and finalize solutions. This balance helps you build problem-solving skills and technical credibility on the job.

When to use it

• Drafting code snippets, architecture ideas, or documentation outlines before reviewing and adapting them.
• Troubleshooting complex bugs where AI suggestions need your context and experience to validate.
• Writing proposals or emails where AI drafts can speed up work but must reflect your accurate intent.
• Learning new tech topics by cross-checking AI explanations with official docs or trusted sources.

The move, step by step

1. Start by asking ChatGPT for a rough draft or framework of the solution you want — for example, boilerplate code or a checklist.

2. Carefully review the AI output line by line. Think about the assumptions ChatGPT made and whether they fit your project context.

3. Cross-reference suggested approaches with official documentation, internal standards, or your own knowledge. Example: Check AWS SDK API calls on AWS docs if ChatGPT generated them.

4. Modify the chatbot’s draft to fix inaccuracies, handle edge cases, or improve efficiency based on what you learn.

5. Test the final version yourself. Run unit tests, integration tests, or manual checks — don’t just trust the AI.

6. Reflect on what you learned by comparing AI suggestions and your corrections. Save notes or snippets for future reference.

7. Over time, reduce dependency on ChatGPT for answers you can confidently generate yourself.

Example

Input:
“Generate Python code to list objects in an AWS S3 bucket.”

ChatGPT output:

import boto3

s3 = boto3.client('s3')
response = s3.list_objects_v2(Bucket='my-bucket')
for obj in response.get('Contents', []):
    print(obj['Key'])

Your verification & improvement:

• Confirmed method list_objects_v2 exists on boto3 client (AWS docs).
• Added error handling for empty buckets and exceptions:

import boto3
from botocore.exceptions import ClientError

s3 = boto3.client('s3')

try:
    response = s3.list_objects_v2(Bucket='my-bucket')
    for obj in response.get('Contents', []):
        print(obj['Key'])
except ClientError as e:
    print(f"Error accessing bucket: {e}")

Common mistakes

• Mistake: Accepting ChatGPT output without verification → Fix: Always check official docs or your own knowledge.
• Mistake: Using AI suggestions verbatim in production → Fix: Test and adapt code to your environment.
• Mistake: Losing context by relying entirely on AI → Fix: Study underlying systems to understand what AI proposes.
• Mistake: Skipping error handling because AI didn’t include it → Fix: Add appropriate safeguards manually.
• Mistake: Ignoring learning opportunities in AI drafts → Fix: Compare AI answers to your experience to grow your skills.

Next step

Take a recent piece of code or task where you used ChatGPT. Follow the steps above to review, verify, and improve the AI-generated content with your own expertise. Then come back and try the next move from the video.