ProblemWell covered5 sources

How should a software engineering workflow change for AI?

The tools changed faster than the practices around them. The sources describe what stays (design, review, ownership) and what moves: where the typing happens, and what a senior engineer now spends the day on.

Everything on this page is AskNex's own paraphrase of what the named sources argue, with each idea attributed to its book. Nothing is quoted, and no disagreement is shown unless the corpus records one. This page orients; it does not answer your situation.

In these topics
01Part

The ideas involved

The concepts the corpus connects to this problem, most central first. Each is source-independent; the books argue about it below.

  1. 01

    AI-assisted software engineering

    Specifying, steering and verifying code produced by a model, rather than typing it.

  2. 02

    Working with AI

    Using a fast, fallible model as a collaborator whose strengths are uneven and must be mapped.

  3. 03

    Software quality and review

    Correctness, maintainability and safety of software, which a fast first draft does not supply.

  4. 04

    Career capital

    Rare and valuable skills bought with sustained effort, then spent on autonomy.

  5. 05

    Validated learning

    Treating a venture as a series of tests rather than a plan to execute.

02Part

What the sources argue

The best-attested position from each of the most relevant books, in AskNex's words.

  1. 01

    The scarce engineering skill becomes specifying, steering and verifying; engineers who delegate well multiply their throughput.

    ArguesAI-assisted software engineering
    Kim & Yegge, Vibe Coding
  2. 02

    Use the model for boilerplate, exploration and explanation; keep architecture and trade-offs human.

    ArguesAI-assisted software engineering
    Osmani, Beyond Vibe Coding
  3. 03

    Bring the model into every task once, so you learn where it earns a place and where it fails.

    ArguesWorking with AI
    Mollick, Co-Intelligence
  4. 04

    Computers are complements to people, not substitutes; the valuable businesses pair human judgement with machine scale.

    ArguesWorking with AI
    Thiel & Masters, Zero to One
  5. 05

    Autonomy is bought with rare and valuable skill, not chosen by following passion.

    ArguesCareer capital
    Newport, So Good They Can't Ignore You
03Part

Where they converge

Ideas on which two or more books make a claim. Agreement here is attested, not assumed.

AI-assisted software engineering

2 sources converge

Specifying, steering and verifying code produced by a model, rather than typing it.

  • The scarce engineering skill becomes specifying, steering and verifying; engineers who delegate well multiply their throughput.
    Kim & Yegge, Vibe Coding
  • Use the model for boilerplate, exploration and explanation; keep architecture and trade-offs human.
    Osmani, Beyond Vibe Coding

Working with AI

3 sources converge

Using a fast, fallible model as a collaborator whose strengths are uneven and must be mapped.

  • Bring the model into every task once, so you learn where it earns a place and where it fails.
    Mollick, Co-Intelligence
  • Computers are complements to people, not substitutes; the valuable businesses pair human judgement with machine scale.
    Thiel & Masters, Zero to One
  • Refusing these tools out of pride or fear now costs more than learning them badly would.
    Kim & Yegge, Vibe Coding

Software quality and review

2 sources converge

Correctness, maintainability and safety of software, which a fast first draft does not supply.

  • The model gets you most of the way quickly; the last stretch of edge cases, integration and correctness is still engineering.
    Osmani, Beyond Vibe Coding
  • Vibe coding accelerates exploration; production software still needs tests, review and architecture the model cannot own.
    Kim & Yegge, Vibe Coding

Career capital

3 sources converge

Rare and valuable skills bought with sustained effort, then spent on autonomy.

  • Autonomy is bought with rare and valuable skill, not chosen by following passion.
    Newport, So Good They Can't Ignore You
  • Expertise is what lets you judge the output; skipping the apprenticeship to rely on the model leaves you unable to evaluate it.
    Mollick, Co-Intelligence
  • Pursue knowledge that cannot be trained for, then attach leverage to it.
    Jorgenson, The Almanack of Naval Ravikant

Validated learning

2 sources converge

Treating a venture as a series of tests rather than a plan to execute.

  • Treat the plan as a set of assumptions and test the riskiest one first.
    Ries, The Lean Startup
  • Iteration without a definite thesis produces copies, not new value.
    Thiel & Masters, Zero to One
04Part

Where they disagree

Only tensions the corpus records, with the reviewed resolution when there is one.

Thiel & Masters against Ries

Tension 1Validated learning
Agrees
Iteration without a definite thesis produces copies, not new value.
Thiel & Masters, Zero to One
Disagrees
Treat the plan as a set of assumptions and test the riskiest one first.
Ries, The Lean Startup
Verdict
Test cheaply where the market is knowable; commit to a thesis where it is not yet visible.

Mollick qualified by Kim & Yegge

Tension 2Career capital
Agrees
Expertise is what lets you judge the output; skipping the apprenticeship to rely on the model leaves you unable to evaluate it.
Mollick, Co-Intelligence
Disagrees
Refusing these tools out of pride or fear now costs more than learning them badly would.
Kim & Yegge, Vibe Coding
Verdict
Adopt the tools now, but not in place of the apprenticeship: use them where you can already judge the output, and build the fundamentals unaided where you cannot.

Osmani qualified by Kim & Yegge

Tension 3Software quality and review
Agrees
The model gets you most of the way quickly; the last stretch of edge cases, integration and correctness is still engineering.
Osmani, Beyond Vibe Coding
Disagrees
The scarce engineering skill becomes specifying, steering and verifying; engineers who delegate well multiply their throughput.
Kim & Yegge, Vibe Coding
Verdict
The leverage is real, and so is the last stretch: delegate the draft, keep the engineering judgement, and budget time for the part the model cannot finish.
05Part

Where to start

One practical move from each of the most relevant books. Places to begin, not a plan; the plan is what a personalised brief writes.

  1. 01Tighten the loop

    Work in small, reviewable increments; never accept a change larger than you can read.

    From Vibe Coding
  2. 02Explain before merge

    Write one paragraph on what a generated change does and why. If you cannot, do not merge it.

    From Beyond Vibe Coding
  3. 03Map the frontier

    For one week, run every task past the model once and record where it helped, hurt or hallucinated.

    From Co-Intelligence
07Part

What this page can't tell you

AskNex would rather say less than imply more.

  • This page orients; it does not know your situation. The personalised brief applies these sources to what you actually describe.
09Apply this to me

This page knows the sources.
It doesn't know you.

Describe your actual situation and AskNex writes a brief from these same sources: the short answer, where they agree and disagree for your case, and what to do this week.

Starts from

How should a software engineering workflow change for AI?

Free to ask. The brief is generated only when you run it, and kept to your account if you sign in.