Module 1 Quiz: The AI-Augmented Engineer

Six questions. Click an option to check it.

---

Question 1

Which of the following got more expensive, not cheaper, when AI coding assistants arrived?

1. Writing boilerplate and adapters
2. Reviewing code and knowing precisely what you want
3. Translating an OpenAPI spec into a client
4. Reading an unfamiliar module

---

Question 2

Which responsibility can be delegated to an assistant?

1. Deciding what to build and what is out of scope
2. Being the final judge of whether the result is correct
3. Owning the outcome in a post-mortem
4. Drafting a first version of the acceptance criteria

---

Question 3

You asked for a one-line bug fix and received a diff touching three files, renaming an interface and adding a dependency. Which failure mode is this?

1. Confident fabrication
2. Scope drift
3. Local optimum
4. Context overflow

---

Question 4

In the five-step delivery loop, where does a failed verification send you?

1. Back to generation, with a request to fix the bug
2. Back to the specification, because the first suspect is an ambiguous description
3. Forward to release behind a feature flag
4. Back to intent, every time

---

Question 5

Which of these is a verification in the sense used in this module?

1. Asking the same assistant "is this correct?"
2. Reading the assistant's explanation of its own diff
3. Running the test suite that was written before the code
4. Noticing that the code looks idiomatic

---

Question 6

The assistant made the failing tests pass. What should you check first?

1. That the commit message is descriptive
2. That the test files were not modified
3. That the code uses the latest language features
4. That the diff is at least 100 lines