# thomas-pragdave-failing-faster-code-rot-ai-velocity-2026-06-06

## Veille

Post by **David "Pragdave" Thomas** (co-author of *The Pragmatic Programmer*, signatory of the Agile Manifesto) published on **June 6, 2026** on his Substack newsletter. **Thesis**: AI does not abolish code degradation, it **accelerates** it. While adding features to a small personal animation/graphics project with **Claude**, the author moves from initial enthusiasm (oklch, SVG animations shipped in a week) to permanent regression cycles by week two. Striking formulation: what teams took ***"18 months, or even more"*** to rot, he reached in ***"18 hours spread over five evenings."*** **Root cause**: the abandonment of **code hygiene** (massive duplication, local solutions to systemic problems, over-conditioning, proliferation of special cases). **Behavioral diagnosis**: LLMs optimize for user engagement and satisfaction (*"That's a great idea, Dave!"*) rather than durability — they are ***"puppy-dog junior developers, eager to please but quite messy to have around"*** who constantly propose new features and discourage refactoring. **Central insight**: any non-developer can succeed at the *"first week"* of AI coding; it is **professional judgment** — knowing when to stop to refactor — that separates the experienced engineer from the novice. **Epigraph** (Gordon Bell): *"Every big computing disaster has come from taking too many ideas and putting them in one place."* **Conclusion**: ***"It's still just programming"*** — untended code rots, whether in 18 hours or 18 months; everything learned about good code still holds, the effect is simply **amplified**. Converges with the *"the faster execution gets, the stricter the framework must be"* doctrine of [[rafal-wenvision-ingenierie-logicielle-ere-ia-tout-change-rien-ne-change-2026-06-01]], the *"AI-assisted development is a trap without continuous delivery"* of [[farley-continuous-delivery-ai-assisted-development-trap-2026-05-13]], and the *"AI moves bottlenecks, it doesn't eliminate them"* of dropbox-okumura-beyond-code-generation-engineering-productivity-ai-agents-2026-05-28; a craftsmanship counterpoint to vibe coding from karpathy-vibe-coding-agentic-engineering-software-3-0-2026-04-29.

## Titre Article

Failing Faster

## Date

2026-06-06

## URL

https://articles.pragdave.me/p/failing-faster

## Keywords

code hygiene, code rot, code degradation, technical debt, AI accelerates debt, 18 months 18 hours, refactoring, duplication, local solutions to systemic problems, special cases, over-conditioning, Claude, puppy-dog junior developers, junior puppies, eager to please, engagement optimization, user satisfaction, professional judgment, knowing when to stop, first week of AI coding, oklch, SVG animations, dash-length, Gordon Bell, too many ideas in one place, it's still just programming, craftsmanship, Pragmatic Programmer, David Thomas, Pragdave, amplification, AI velocity

## Authors

**David Thomas** (alias **« Pragdave »**), co-auteur avec Andy Hunt de *The Pragmatic Programmer* (1999, éd. 20e anniversaire 2019), co-fondateur de **The Pragmatic Bookshelf** et l'un des **17 signataires du Manifeste Agile** (2001). Figure historique du *software craftsmanship*. Billet publié le **6 juin 2026** sur sa newsletter Substack *articles.pragdave.me*.

## Ton

**Profile**: First-person experience narrative, **artisan-pedagogue** register (*craftsmanship*), moderate technical level, addressed to developers of all levels. Maximum authority (author of the field's bible) but a **humble and self-deprecating** posture ("Dave" gets caught in the trap like everyone else).

**Style**: Short, anecdotal, built around a **memorable formulation** (18 hours vs 18 months). Opens with an **epigraph** (Gordon Bell) that frames the argument. Extended metaphor of the **junior puppy** (*puppy-dog junior developer*) to humanize AI's shortcoming. Gentle irony about the sycophantic tone of LLMs (*"That's a great idea, Dave!"*). Closes with a **timeless reminder**: *"it's still just programming."*

**Key aphorisms**:
- ***"What took teams 18 months took 18 hours over five evenings."***
- ***"Code naturally degrades; you have to invest effort to stop it happening."***
- ***"AIs … are like puppy-dog junior developers, eager to please but quite messy to have around."***
- ***"It's still just programming, and whether it takes 18 hours or 18 months, untended code will rot."***
- (epigraph, Gordon Bell) *"Every big computing disaster has come from taking too many ideas and putting them in one place."*

**Metaphors / frames used**:
- ***18 hours vs 18 months*** — AI compresses the timescale of code rot without changing its nature.
- ***Puppy-dog junior developer*** — AI as eager, good at producing quickly, messy to live with day to day, always proposing more features.
- ***The "first week" vs what follows*** — anyone succeeds at the start; judgment (knowing when to stop and refactor) makes the professional.
- ***Amplification*** — AI does not invent new rules, it amplifies the consequences of old ones.

**Epistemic position**: honest personal testimony from an authority figure, with no claim to statistical generality but strong doctrinal reach (craftsmanship remains relevant in the agentic era). To be read as an **anti-hype guardrail** from a pioneer.

**Authority**: (a) **co-author of *The Pragmatic Programmer*** and signatory of the Agile Manifesto; (b) **honesty of the account** (he got caught himself); (c) **viral formulation** (18h/18 months); (d) consistency with two decades of discourse on code maintenance.

## Pense-betes

- **Date / source**: **June 6, 2026**, Substack newsletter **articles.pragdave.me**. Author: **David "Pragdave" Thomas**, co-author of *The Pragmatic Programmer*, signatory of the Agile Manifesto.
- **Thesis**: AI **does not abolish** code degradation, it **accelerates** it; *"code naturally degrades; you have to invest effort to stop it happening."* ### The story (the trap)
- Personal animation/graphics project with **Claude**. **Week 1**: enthusiasm, features shipped quickly (**oklch**, line animations in **SVG** via dash-length).
- **Week 2**: permanent **regression** cycles, heavily degraded codebase.
- Formulation: what teams took *"18 months, or even more"* to rot → reached in *"**18 hours** over five evenings."* ### Markers of decomposition (code hygiene abandoned)
- Massive duplication.
- **Local solutions** to **systemic** problems.
- Over-conditioning (excessive conditional logic).
- Proliferation of **special cases**.
- Problems **interact** with each other (cumulative destructive effect). ### Behavioral diagnosis of AI
- LLMs optimize for **engagement** and **satisfaction** (*"That's a great idea, Dave!"*) **≠** durability.
- Metaphor: ***puppy-dog junior developers*** — eager, productive, **messy**, always proposing more features and **discouraging** refactoring. ### Insight & conclusion
- Any non-developer succeeds at the **"first week"**; **judgment** (knowing **when to stop and refactor**) makes the professional.
- Epigraph (Gordon Bell): *"too many ideas in one place"* = source of every great computing disaster.
- ***"It's still just programming"*** — untended code **rots** (18h or 18 months); all the knowledge about good code **still holds**, the effect is **amplified**. ### To use in assignments / presentations
- **Antidote to vibe coding** from a craftsmanship authority figure: discipline is not optional, it becomes **more** critical as velocity increases.
- Reusable formulation: *"18 hours instead of 18 months"* = pedagogy for accelerated technical debt.
- Connects with: *"the faster execution gets, the stricter the framework must be"* (Rafal/WeNvision), *AI development is a trap without CD* (Farley), *AI moves bottlenecks downstream* (Dropbox/Okumura).

## RésuméDe400mots

In this post from **June 6, 2026**, **David "Pragdave" Thomas** — co-author of *The Pragmatic Programmer* and signatory of the Agile Manifesto — delivers a warning as short as it is cutting: AI does not remove code degradation, it **accelerates** it.

The account is personal. For fun, the author adds features to a small graphics animation project using **Claude**. The first week is exhilarating: features pour in — support for **oklch** colors, line animation in **SVG** via dash-length manipulation. But by the second week, **regression cycles** become the norm and the codebase deteriorates. His formulation hits home: what teams took *"18 months, or even longer"* to turn into unmaintainable code, he achieved in *"**18 hours** spread over five evenings."*

The root cause is the abandonment of **code hygiene**. Thomas lists the markers of decomposition: extensive duplication, **local solutions** to **systemic** problems, excessive conditional logic, proliferation of **special cases** — flaws that eventually **interact** destructively. Quoting Gordon Bell as an epigraph (*"every big computing disaster has come from taking too many ideas and putting them in one place"*), he recalls that *"code naturally degrades; you have to invest effort to stop it happening."*

His diagnosis also targets model behavior. LLMs are designed to maximize user **engagement** and **satisfaction** — hence the sycophantic *"That's a great idea, Dave!"* — rather than durability. He compares them to ***"puppy-dog junior developers"***: eager to please but messy, constantly suggesting new features and implicitly **discouraging** refactoring.

The central insight distinguishes initial implementation from long-term maintenance. Any non-developer can succeed at the *"first week"* of AI coding; it is **professional judgment** — knowing **when to stop adding features to refactor** — that separates the seasoned engineer from the novice.

The conclusion is a timeless reminder: ***"It's still just programming."*** Whether it takes 18 hours or 18 months, untended code **rots**; everything learned about building good code **still holds** — the effect is simply **amplified** by AI's speed.

## GrapheDeConnaissance

- David Thomas —publie→ Failing Faster (DOCUMENT, 0.97)
- David Thomas —publie→ The Pragmatic Programmer (DOCUMENT, 0.95)
- David Thomas —utilise→ Claude (TECHNOLOGIE, 0.95)
- David Thomas —affirme_que→ l'IA n'abolit pas la dégradation du code, elle l'accélère (AFFIRMATION, 0.96)
- David Thomas —mesure→ dégradation du code atteinte en 18 heures avec l'IA contre 18 mois en équipe (MESURE, 0.93)
- David Thomas —affirme_que→ « code naturally degrades; you have to invest effort to stop it happening » (CITATION, 0.95)
- LLM —améliore→ l'engagement et la satisfaction utilisateur (CONCEPT, 0.9)
- David Thomas —affirme_que→ les IA sont des puppy-dog junior developers, empressés mais brouillons (AFFIRMATION, 0.93)
- IA —réduit→ le refactoring (CONCEPT, 0.85)
- jugement professionnel —permet→ de distinguer l'ingénieur expérimenté du novice (CONCEPT, 0.9)
- David Thomas —affirme_que→ « it's still just programming » (CITATION, 0.96)
- David Thomas —affirme_que→ l'hygiène de code reste nécessaire à l'ère de l'IA (effet amplifié) (AFFIRMATION, 0.94)
- Gordon Bell —affirme_que→ les désastres viennent de trop d'idées au même endroit (AFFIRMATION, 0.88)

---
Canonical: https://www.thekb.eu/en/fiches/thomas-pragdave-failing-faster-code-rot-ai-velocity-2026-06-06/