AI in Automated Code Migration Quiz

Answers and explanations appear as you take the quiz.

1. Why is AI-assisted code migration valuable?
   • Manual migration is slow, costly, and error-prone, and AI can rewrite at scale
   • It removes the need for any testing
   • It makes programs run faster automatically
   • It eliminates the need for programming languages
2. What is the typical safe workflow for AI code migration?
   • The AI commits changes directly with no review
   • Human-in-the-loop: AI proposes diffs, tests verify, engineers approve
   • Engineers rewrite everything and ignore the AI
   • The AI deletes the old code immediately
3. How do migration tools reduce hallucinated or wrong APIs?
   • By trusting the model's first output completely
   • By disabling all type checking
   • By compiling the code and running existing tests to validate edits
   • By converting code to images
4. Why pair LLMs with AST parsing and static analysis?
   • To make the code look prettier
   • To translate comments into other human languages
   • To compress the source files
   • To precisely locate what must change and give the model scoped context
5. Which of these is a classic real-world migration target?
   • Moving decades-old COBOL banking systems to modern languages
   • Converting JPEG images to PNG
   • Renaming files in a folder
   • Changing a website's color scheme
6. As AI migration tools improve, where does the main bottleneck shift?
   • To writing the changes by hand
   • To reviewing the AI-generated changes safely
   • To buying faster hardware
   • To learning new human languages
7. In the guide’s verification loop, what should happen when a generated migration edit fails a test?
   • Delete the tests and accept the edit
   • Treat the failure as proof the migration is complete
   • Use the failure as feedback for another edit, then validate again
   • Change unrelated files until the test passes
8. What is the role of AST parsing and static analysis in scoped migration?
   • Automatically decide whether the new code is legally licensed
   • Replace all testing with a language model opinion
   • Turn source files into audio
   • Locate relevant code constructs and provide focused context for transformation