Saltar al contenido principal

Module 1: The AI-Augmented Engineer

Learning objectives​

By the end of this module, you will be able to:

  • Explain, in one sentence, what an AI coding assistant is good at and what it is structurally unable to do.
  • List the three responsibilities that remain with the engineer regardless of tooling.
  • Describe the five-step delivery loop (intent, specification, generation, verification, release) and say which steps the assistant owns and which steps you own.
  • Recognise the three most common failure modes of AI-generated code and the check that catches each one.

Module content​

Chapter 1: What changes, what does not​

The honest inventory. Which parts of the job got cheaper, which parts got more expensive, and why "the AI wrote it" is never an acceptable answer in a post-mortem.

Chapter 2: The new delivery loop​

From "write code, then test it" to "state intent, specify, generate, verify, ship". Where the human sits in each step, and why verification is now the job rather than an afterthought.

Quiz​

Six questions to check that the vocabulary of this module has landed before you move on.

Why start here​

Most teams adopt an assistant the way they adopt a new keyboard shortcut: individually, informally, and without changing anything else. That works for a few weeks. Then the review queue fills up with large, plausible, half-understood pull requests, and the team concludes that "AI does not work for us". It usually does work. What did not happen is the change of workflow that this module describes.