Do Not Guess At Performance

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-coder-performance-profile-creators

1. A developer notices their web app feels sluggish. What should they measure before asking an AI assistant to optimize the code?

   1. How many times the word 'performance' appears in their code
   2. The app's color scheme and font choices
   3. Render time, network time, query time, and bundle size
   4. The number of lines in their JavaScript files

2. Why is it important to profile an application before requesting AI-powered optimizations?

   1. Profiling gives the AI a specific target rather than a vague impression of slowness
   2. Profiling makes the code look more impressive to other developers
   3. Profiling automatically fixes the performance issues
   4. Profiling is required by most web hosting platforms

3. What does the principle 'name the job before naming the tool' advise developers to do?

   1. Choose a programming language before starting a project
   2. Write the tool's name in comments within the code
   3. Always use the most popular AI coding assistant available
   4. Define the specific problem or task before specifying which AI tool to use

4. When requesting an AI to optimize code, what is meant by 'the smallest useful scope'?

   1. Writing the optimization request in as few words as possible
   2. Creating a task small enough for the AI to complete quickly and usefully
   3. Using the fewest possible lines of code in the prompt
   4. Focusing only on syntax-level changes, not logic changes

5. How should you evaluate the result of an AI-generated performance optimization?

   1. Compare it to the AI's original prompt to ensure they match
   2. Only check if the code compiles without errors
   3. Run it as a user would, testing the actual user experience
   4. Run it as a fan of the tool, celebrating whatever it produces

6. Before sharing AI-generated code with a team, what three things should a developer inspect?

   1. The color scheme, layout, and formatting
   2. The file size, creation date, and author
   3. The variable names, comment style, and indentation
   4. The diff, data access patterns, and failure path

7. Why do AI-generated optimizations often chase the wrong bottleneck?

   1. AI can only optimize code, not hardware
   2. AI models intentionally make code slower
   3. AI lacks real-time data about the actual application's performance characteristics
   4. AI always chooses the most complex solution available

8. What does profiling provide that a 'vibe' or feeling about performance does not?

   1. Concrete, measurable data about where time is actually spent
   2. A list of all the functions in the application
   3. A score that can be shared on social media
   4. A way to complain about slow code

9. When evaluating AI-generated code, which question addresses security concerns?

   1. What should the user be able to do when finished?
   2. What test proves the change works?
   3. What rollback path exists if the output is wrong?
   4. What data should the app or agent never expose?

10. What makes code 'observable' in the context of AI-assisted development?

    1. The code has extensive comments explaining each line
    2. The code uses many console.log statements
    3. The code is available in a public repository
    4. The code's behavior and performance can be measured and monitored

11. Why is having a rollback path important when using AI to generate code changes?

    1. It allows reverting to a working state if the AI-generated change causes problems
    2. It makes the code look more professional
    3. It is required by most version control systems
    4. It increases the bundle size of the application

12. What does 'bundle size' refer to in web performance?

    1. The total size of all physical files in a web server's folder
    2. The combined file size of JavaScript and other assets downloaded by a user's browser
    3. The number of separate files in a project
    4. The physical dimensions of images used on a website

13. In performance profiling, what is 'render time' measuring?

    1. How long the code takes to compile before running
    2. How long a user waits for a file download to complete
    3. How long it takes the server to process a database query
    4. How long the browser takes to display content on the screen

14. Why should a developer inspect the 'failure path' of AI-generated code?

    1. To compare the failure path to the success path
    2. To understand what happens when errors occur and prevent security issues
    3. To ensure the code fails as quickly as possible
    4. To count how many error messages the code produces

15. What separates a working demo from production-ready code, according to this lesson?

    1. Whether the code was written by an AI or a human
    2. Whether the code passes all automated tests
    3. Whether the code uses modern JavaScript syntax
    4. Whether the code is observable, reversible, and safe for others to use