14 Reconstructing Work
Vibe coding began as a story about coding, but it ends as a story about work. The day of a person working with AI, the actual shifts in the job landscape, and even the question “if AI does everything, what am I here to do?” We start not from prediction but from observation of changes that have already happened.
14.1 A Day Working with AI
Pre-draft. Check ownership and status in TOPICS.md at the repository root.
Brief. The day of a practitioner who uses AI deeply is being restructured around time spent delegating, reviewing, and running multiple tasks in parallel, rather than time spent typing code directly. This section starts from observation, not prediction. The core method of this section is to directly interview people who work deeply with AI and record, as primary source material, how their day has changed. This corresponds to the introduction of Chapter 14 (Reconstructing Work).
Seed questions
- If you reconstruct the daily routine of a developer or researcher who uses AI deeply hour by hour, what stands out as having changed?
- How does the cycle of delegation and review actually play out, and has the time spent reviewing increased or decreased relative to the time spent generating?
- Does the practice of running multiple agents in parallel actually increase productivity, or does it increase the burden of management?
- What tasks remain that people insist on doing themselves rather than delegating to AI during the day, and what is the criterion for that?
- How does this change appear differently depending on occupation (developers, researchers, founders)?
Investigation pointers
- Directly recruit practitioners and researchers who use AI deeply and conduct in-depth interviews. The core material for this section is the interviews
- Diversify interview subjects so that occupation, career experience, and tools used differ
- If there are already published primary-source essays or blogs in the form of a “day in the life” record, refer to them, but do not let them replace the interviews
14.2 The Changing Job Landscape: A Data-Driven Look
Pre-draft. Check TOPICS.md at the repository root for owner and status.
Brief. There are many claims that AI is changing the job landscape for developers, but the quality of the evidence varies widely. This section looks for empirical research on job postings, wages, and employment scale in order to separate exaggerated narratives from changes that are actually confirmed. The only way this book is allowed to cite figures is by directly checking the original data and methodology. This section does not predetermine any specific figures or conclusions. The person in charge must not assert any figure without verification, and should write claims only after verifying them against empirical research and original data.
Seed Questions
- What are reliable data sources that track changes in the number of developer job postings and required skills, and what are their methodologies?
- Can statistics on developer wage trends be causally linked to AI adoption, or do they only show correlation?
- For the hiring-decline figures frequently cited in the media, what is the original source, and is the methodology verifiable?
- How do patterns of change in employment data differ by country and by industry?
- Where empirical studies reach conflicting conclusions, is the cause a difference in methodology or a difference in the observation period?
Research Pointers
- Check labor economics academic papers and official employment statistics (original data published by governments and international organizations) as primary sources.
- For reports that hiring platforms publish themselves, be sure to check and state the limitations of their methodology and sample.
- Use any figure in this section only after directly checking the original data. Do not assert figures that could not be verified; leave them marked as “to be verified.”
14.3 Identity Question: What Kind of Person Am I?
Pre-draft. Check TOPICS.md at the repository root for owner and status.
Brief. If AI writes a substantial portion of the code, how does the person who made that code define what kind of person they are. This piece covers how the self-definition of developers, researchers, and creators is actually changing, and where the old discourse of craftsmanship or mastery goes in the face of this change. The method of this piece is to treat the voices of those involved, not statistics, as the primary source. As the final piece of Chapter 14, it closes the entire book with the identity question.
Seed Questions
- If someone has shifted from “I am a person who writes code” to a different self-definition, how does that new definition (designer, reviewer, orchestrator, etc.) actually get expressed in interviews?
- Is the discourse of craftsmanship and mastery discarded in the face of AI-assisted production, or is it redefined in a different form?
- How do reactions to this identity shift differ by generation and by career stage?
- What grounds do people who continue to hold onto the self-definition of “maker” give for doing so?
- Are there comparable cases to reference from identity debates in fields that experienced tool automation earlier, such as photography or music production?
Research Pointers
- Conduct in-depth interviews directly with developers and creators about identity change. Interviews are the core material for this piece
- Check literature on identity debates in fields that experienced automation earlier, such as photography and music, as comparative material
- Cross-reference with §13.3 (The Ladder of Expertise) but divide roles: this piece covers identity and self-narrative, §13.3 covers learning pathways