# frizzo-linkedin-year-claude-code-output-doubled-attention-span-2026-05-05

## Veille

LinkedIn Pulse op-ed by Alexandre Frizzo after a year of daily use of Claude Code, offering a **nuanced assessment** rare in the 2026 corpus — productivity **multiplied by 3-5×** in his case (consistent with Wescale, and in line with the median of committed practitioners; the elite tail goes much higher, cf. Cherny *few dozen PRs/day + 150 PRs record* and Karpathy *"peaks much higher than 10×"*), but **hidden cognitive costs** acknowledged. Pivot thesis: ***"the new bottleneck is supervision"*** — the job has changed shape, one no longer *writes* code, one *decides* about code generated by agents. Gains: 3-5× output, previously infeasible projects now achievable (yak-shaving, boilerplate), near-zero cost of experimentation. Acknowledged losses: ***"writing muscle"*** atrophied (manual code now feels *effortful*), **rare deep flow state** (constant context-switching between supervisions), **diminished ownership satisfaction** (*"code is good, but isn't quite mine"*). Unresolved tensions: **FOMO** (*"every hour I'm not at the keyboard is an hour an agent could be earning for me"*), **review quality** at 3-5× volume, **skill atrophy**. Statistics cited: median 3-4h effective coding out of an 8h day, **23 min** context recovery per interruption (Gloria Mark study), 15-25 min flow entry, 500% productivity in flow (McKinsey). Exemplary epistemic position: simultaneously rejects the *"AI is bad"* narrative and uncritical enthusiasm. A welcome counterweight to Cherny's *"coding is solved"* (2026-05).

## Titre Article

A Year With Claude Code: My Output Doubled. My Attention Span Didn't.

## Date

2026-05-05

## URL

https://www.linkedin.com/pulse/year-claude-code-my-output-doubled-attention-span-didnt-frizzo-atmoe/

## Keywords

Alexandre Frizzo, LinkedIn Pulse, year with Claude Code, output doubled attention span didn't, productivity 3-5x, supervision bottleneck, job changed shape, write hard parts review easy parts, deciding about code, multi-codebase supervision, writing muscle atrophy, deep flow state rare, ownership satisfaction diminished, code is good isnt quite mine, FOMO overnight agents, review quality at scale, skill atrophy question, yak-shaving boilerplate previously impossible, near-zero experimentation cost, Gloria Mark 23 minutes context recovery, 15-25 minutes flow entry, 15-30 minutes mental context rebuild, McKinsey flow 500% productivity, median 3-4 hours coding per day, every hour not at keyboard is an hour agent could be earning, AI is bad rejected, uncritical enthusiasm rejected, honest practitioner assessment, cognitive costs hidden, traditional deep work irrelevant, quality control at scale, weekend evening work pressure

## Authors

Alexandre Frizzo (auteur LinkedIn Pulse, identité tech non précisée par le post au-delà du nom — auteur d'une tribune one-year retrospective Claude Code).

## Ton

**Profile**: LinkedIn Pulse op-ed in *one-year retrospective* format signed by a practitioner (developer or tech lead) after twelve months of daily use of Claude Code. A hybrid format between *honest journal* and strategic essay, reflective rather than prescriptive tone. Target audience: fellow developers on LinkedIn (the French- and English-speaking tech professional network), tech leads seeking to balance the AI discourse, corporate audiences fatigued by X10 promises. Secondary audience: managers and CHROs confronted with the human cost of AI transformation.

**Style**: First-person voice, clear English, conversational-confessional register. Frizzo alternates between **quantified observation** (*"3-5x productivity multiplier"*, *"23 minutes per interruption"*) and **frank introspection** (*"the code is good, but isn't quite mine"*). No hype, no catastrophism — this is the ethic of the *practitioner-journalist* reporting what he lives.

**Key aphorisms**:
- ***"My output doubled. My attention span didn't."*** (title — a perfectly symmetrical formula that condenses the thesis).
- ***"The new bottleneck is supervision."***
- ***"Code is good, but isn't quite mine."***
- ***"Every hour I'm not at the keyboard is an hour an agent could be earning for me."***

**Elaborated metaphors**:
- ***"Writing muscle"*** — a bodily metaphor that makes cognitive atrophy tangible (*manual coding feels effortful now*).
- **Bottleneck shift** — the concept becomes pivotal: protecting *deep work state* becomes *irrelevant* when generation is fast; *quality control at scale* requires different defenses.
- **Supervision** vs **creation** — the canonical opposition of the transformed job.

**Epistemic position**: exemplary for 2026. Frizzo explicitly rejects the **two available narratives** (*"AI is bad" / uncritical enthusiasm*) and holds an **honest third position** — real gains + real costs. He raises **unresolved tensions** rather than conclusions, which is rare in typically *advice-driven* LinkedIn content. This is **authored prose**, not an influencer thread.

**Authority**: built through (a) **length of use** (one year, daily) giving rare longitudinal experience, (b) **methodological honesty** (citing studies — Gloria Mark, McKinsey — to support his introspective observations), (c) **refusal of clickbait** despite a brilliant title, (d) **precision of figures** (3-5×, 23 min, 15-25 min, 500%) showing an author who did his research.

## Pense-betes

- **Date / source**: 5 May 2026, **LinkedIn Pulse**. Author: **Alexandre Frizzo**. URL: https://www.linkedin.com/pulse/year-claude-code-my-output-doubled-attention-span-didnt-frizzo-atmoe/
- **Format**: *one-year retrospective* — a year of daily use of Claude Code, an honest assessment.
- **Title-aphorism**: ***"My output doubled. My attention span didn't."*** — a perfect symmetry that condenses the diagnosis.
- **Central thesis**: ***"The new bottleneck is supervision. Reading what an agent produced, deciding if it's right, integrating it, catching the subtle things it got wrong."*** ### The job has changed shape
- **Before**: *"write hard parts, review easy parts"* (the developer writes the difficult logic and reviews the details).
- **After**: *"deciding about code"* across **multiple simultaneous codebases** — the developer has become a supervisor of agents.
- **The bottleneck has shifted**:
- **Old bottleneck**: protecting *deep work state* (entering flow, deep concentration).
- **New bottleneck**: **high-speed supervision** (reading agent output, deciding correctness, integrating, catching subtleties).
- *"The traditional focus on protecting deep work state became irrelevant once code generation became fast."*
- *"Quality control at scale now requires entirely different defenses."* ### Acknowledged gains
- **3-5× productivity** on a typical day. This figure is **consistent with Wescale X3-X4** (2026-05-03) and reflects the experience of a practitioner committed for 1 year. Not to be confused with the elite tail, which goes much higher: **Cherny** (2026-05) states *"a few dozen PRs/day, 150 PRs in a single day record"* (clearly >10× the usual baseline), **Karpathy** (2026-04-29) specifies *"10× is not the speed up — people who are very good at this peak a lot more than 10×"*, **Curran/Intercom** (2026-04-16) measures that the **top 5% of performers generate 6× the median PR throughput**. Correct reading: **uneven distribution**, 3-5× for the committed median, 10×+ for the elite tail. The myth to debunk is not 10× itself but *"automatic 10× for everyone with a simple prompt"*.
- **Previously infeasible projects now achievable**: yak-shaving, boilerplate, marginal automations — the *long-tail* becomes addressable (consistent with **Wescale 2026-05-03** *"historical needs that remained too costly can finally be addressed"*).
- **Near-zero cost of experimentation** — the mental barrier to trying things is lifted. ### Acknowledged losses
- **"Writing muscle"** atrophied — *"manual coding feels effortful now"*. A bodily metaphor that makes **degradation through disuse** tangible.
- **Rare deep flow state** — *"constant context-switching"* between multiple supervisions. The cognitive flow state protected by 30 years of *Deep Work* literature (Cal Newport) becomes inaccessible.
- **Diminished ownership satisfaction** — ***"the code is good, but isn't quite mine."*** The psychological sense of creation is eroded. ### Statistics cited (supporting evidence) | Measure | Value | Source | |--------|--------|--------| | Median effective coding time out of an 8h day | **3-4 hours** | Developer research | | Context recovery per interruption | **23 minutes** | Classic Gloria Mark study (UC Irvine) | | Entering flow state | **15-25 minutes** | Flow research | | Rebuilding mental context | **15-30 minutes** | Cognition research | | Productivity multiplier in flow | **500%** | McKinsey senior executives study | ### Unresolved tensions (the questions he does not settle) 1. **FOMO Problem** — knowing that Claude can work overnight creates **psychological pressure**. *"Every hour I'm not at the keyboard is an hour an agent could be earning for me."* Weekend/evening work becomes tempting **even when projects are finished**. An inversion of the deep-work logic: it is no longer the absence of productivity that weighs, it is the absence of **active supervision**. 2. **Review Quality at 3-5× Volume** — reviewing at 3-5× the volume risks *skimming*. The industry's historical quality practices assumed a **human pace**. At this throughput level, *subtle bugs* can slip through review. 3. **Skill Atrophy** — Frizzo poses the **open question**: *"does the writing muscle still matter, or is it becoming commoditized?"* He does not settle it — it is a political point. ### Ethical-epistemic position
- **Simultaneous refusal of the two available narratives**:
- *"AI is bad"* — Frizzo explicitly rejects it.
- *Uncritical enthusiasm* — he rejects this too.
- **Third-position stance**: real gains + real costs, **unresolved tensions** rather than conclusions. ### Connection to the watch corpus
- **A welcome counterweight to Cherny's *"coding is solved"*** (2026-05, fiche `cherny-sequoia-coding-is-solved-loops-printing-press-2026-05.md`). Cherny declares 100% of code generated, *"loops are the future"*, an iOS setup with 5-10 sessions / a few hundred agents. Frizzo, with **the same intensive daily use over a year**, concludes that **the cognitive costs are real**. The 2026 corpus needs both voices to stay honest.
- **Numerical convergence — uneven distribution**:
- **Committed median 3-5×** (the most widely documented experience):
- **Frizzo**: 3-5× productivity multiplier over 1 year of daily use.
- **Wescale Usine Logicielle Augmentée** (2026-05-03): *"realistic X3-X4"*.
- **Curran/Intercom** (2026-04-16): 3× R&D productivity in 16 months (entire organization, 500 R&D).
- **DORA Report 2025** (2025-09-23) and **Stanford Denisov-Blanch** (2025-11-23): *quantify AI ROI*.
- **Elite tail 10×+** (practitioners who have entirely reorganized their work):
- **Cherny** (2026-05): *"a few dozen PRs/day, 150 PRs in a single day record"* (clearly >10× a dev's usual baseline).
- **Curran/Intercom top 5%**: **6× the median PR throughput** within an organization already averaging 3× — i.e. **18× the pre-AI baseline** for this subset.
- **Karpathy** (2026-04-29): *"10× is not the speed up — people who are very good at this peak a lot more than 10×"*.
- → **Correct reading**: **uneven** distribution, 3-5× for the committed median, 10×+ for the elite tail. The myth to debunk is *"automatic X10 for everyone"*, not X10 itself. Frizzo and Cherny do not contradict each other — they measure **two different points** of the distribution.
- **Convergent cognitive tensions**:
- **BCG/HBR Brain Fry** (Bedard et al., 2026-03-05): 14% AI brain fry, peak 3 tools, +39% major errors, +39% intent to leave. Frizzo provides the **individual testimony** that the BCG study measures at the population level.
- **Anthropic January 2026** study (cited by Sun NYT 2026-04-30): *"junior engineers using AI agents understood their work less when quizzed afterward."* Frizzo confirms individually: *writing muscle atrophy*.
- **Karpathy *"outsource thinking but not understanding"*** (2026-04-29): Frizzo experiences the tension *"code is good, but isn't quite mine"* which is exactly the loss of *understanding* while *thinking* has been outsourced.
- **Beck *Starving Genies*** (2026-04-03): voluntary scarcity vs FOMO from 24/7 agents. Frizzo identifies FOMO as an **unresolved tension**.
- **Soto *Developer Taste*** (2026-04): taste as the last skill. Frizzo poses *review quality at scale* as a challenge to taste at high throughput.
- **Mornati *What is a Developer When We Use Coding Agents?*** (2026-03-14): 1-day experience, the 70% problem. Frizzo confirms after 1 year: the nature of the job has structurally changed.
- **Klaassen *Stop Coding and Start Planning*** (2025-11-06): compounding engineering, three fidelities. Frizzo lives this transition (planning > coding) on a daily basis.
- **Convergence on "supervision is the new bottleneck"**:
- **Wescale** (2026-05-03): *Strategic Judge* + *Agent Manager* — the human becomes a supervisor.
- **Karpathy** (2026-04-29): agents as *interns with recall but no taste* — the human in charge of *spec*, *judgment*, *oversight*.
- **Andreessen** (2026-02): *orchestrating 10 bots*.
- **Habert PROJ-AI** (2026-05-05): *"agent directives + Decision Records + five validation dimensions"*.
- **Frizzo**: experiences the supervision bottleneck day to day and names the **costs**.
- **FR vs Anglo-Saxon position**:
- Frizzo writes in English on LinkedIn — not specifically French despite the name.
- His **epistemic honesty** aligns more with **French caution** (Wescale, Habert) than with **American optimism** (Cherny, Curran).
- To be used in FR presentations as a **practitioner testimony** complementing Wescale/consulting-firm figures. ### Limitations to flag
- **Author identity poorly documented**: Alexandre Frizzo — no visible institutional credentials, no cited industrial practice. Authority built through the text's coherence, not through external credentials.
- **Sample n=1**: a year of personal use, not a study. But methodological honesty compensates — he cites external studies (Gloria Mark, McKinsey).
- **No verifiable internal figures** on the 3-5× productivity — it is a **subjective estimate**, not an output measurement (vs Curran/Intercom, which measures PRs).
- **No action roadmap** — the text is diagnostic, not prescriptive. Those seeking recommendations should complement it with Wescale, Habert, Klaassen. ### To be used for
- **Executive committee presentations**: balanced, anti-hype practitioner testimony that qualifies X10 promises without demobilizing.
- **Manager awareness** of the **hidden cognitive costs** of AI transformation (FOMO, deskilling, ownership erosion).
- **Ethical debate** on the transformation of the developer's role into a supervisor.
- **Sourced figures**: 3-5× realistic, 23 min Mark, 500% McKinsey flow — figures to integrate into watch fiches, strategic presentations, internal training.

## RésuméDe400mots

Alexandre Frizzo publishes on LinkedIn Pulse, on 5 May 2026, a *one-year retrospective* op-ed on his daily use of Claude Code. The title condenses the thesis: ***"My output doubled. My attention span didn't."*** A nuanced assessment rare in the 2026 corpus — productivity multiplied by **3-5×**, but **hidden cognitive costs** acknowledged.

**The job has changed shape.** Frizzo no longer writes code, he **makes decisions about code generated by agents**. *"Write hard parts, review easy parts"* becomes multi-codebase supervision. The bottleneck has shifted: protecting *deep work state* has become *irrelevant* now that generation is fast; ***"the new bottleneck is supervision"*** — reading agent output, deciding correctness, integrating, catching subtleties. *"Quality control at scale now requires entirely different defenses."*

**Acknowledged gains**: 3-5× productivity multiplier on a typical day, previously infeasible projects now accessible (yak-shaving, boilerplate, *long-tail*), near-zero cost of experimentation.

**Acknowledged losses**: *writing muscle* atrophied (*"manual coding feels effortful now"*), rare deep flow state due to constant context-switching, diminished ownership satisfaction (***"the code is good, but isn't quite mine"***).

**Statistics cited**: median 3-4h effective coding out of an 8h day, **23 min of context recovery per interruption** (Gloria Mark), 15-25 min flow entry, 500% productivity in flow (McKinsey senior executives study).

**Unresolved tensions** that Frizzo raises without settling: (1) **FOMO** — *"every hour I'm not at the keyboard is an hour an agent could be earning for me"* — the psychological pressure of 24/7 agents; (2) **Review quality** — reviewing at 3-5× the volume risks *skimming*, quality practices assumed a human pace; (3) **Skill atrophy** — *"does the writing muscle still matter, or is it becoming commoditized?"*

**Exemplary epistemic position**: Frizzo simultaneously rejects the two available narratives — *"AI is bad"* and *uncritical enthusiasm*. He holds an honest third position, real gains + real costs, **unresolved tensions** rather than conclusions.

**Connection to the watch corpus**: a welcome complement to **Cherny's *"coding is solved"*** (2026-05) — same intensive daily use, but Frizzo sits on the **committed median** (3-5×) while Cherny and the top 5% Curran/Intercom occupy the **elite tail** (10×+). The two do not contradict each other: they measure **two different points** of the productivity distribution. Numerical convergence on the median with **Wescale** (2026-05-03), **Curran/Intercom** average (2026-04-16), **DORA Report 2025**, **Stanford Denisov-Blanch** (2025-11-23). Cognitive convergence with **BCG Brain Fry** (Bedard et al., 2026-03-05), **Anthropic study on junior engineers deskilling** (cited by Sun NYT 2026-04-30), **Karpathy *outsource thinking but not understanding*** (2026-04-29), **Soto *Developer Taste*** (2026-04). To be used as a balanced practitioner testimony for executive committees, manager awareness, and the ethical debate on the developer's transformation into a supervisor.

## GrapheDeConnaissance

- Alexandre Frizzo —publie→ A Year With Claude Code (DOCUMENT, 0.97)
- Alexandre Frizzo —utilise→ Claude Code (TECHNOLOGIE, 0.95)
- Alexandre Frizzo —affirme_que→ "the new bottleneck is supervision" (CITATION, 0.97)
- Alexandre Frizzo —mesure→ productivité 3-5× sur usage quotidien (MESURE, 0.94)
- Alexandre Frizzo —affirme_que→ le métier de développeur a changé de forme, de l'écriture vers la supervision (AFFIRMATION, 0.95)
- Usage prolongé d'agents —réduit→ Writing muscle atrophy (CONCEPT, 0.93)
- Context-switching multi-supervision —réduit→ deep flow state (CONCEPT, 0.93)
- Code généré par agent —réduit→ ownership satisfaction (CONCEPT, 0.92)
- Gloria Mark —mesure→ 23 min de récupération contexte par interruption (MESURE, 0.97)
- McKinsey —mesure→ 500% productivité en flow state (MESURE, 0.93)
- FOMO d'agents 24/7 —permet→ pression psychologique permanente (CONCEPT, 0.94)
- Volume 3-5× —réduit→ Review quality at scale (CONCEPT, 0.93)
- Alexandre Frizzo —s_oppose_à→ narratif "AI is bad" et uncritical enthusiasm (CONCEPT, 0.96)
- Bilan Frizzo —affine→ déclaration "coding is solved" de Cherny (CONCEPT, 0.92)
- Productivité 3-5× —converge_avec→ Wescale X3-X4, Curran 3× (moyenne), DORA 2025 (CONCEPT, 0.93)
- Distribution productivité IA —mesure→ médiane committée 3-5×, tail élite 10×+ (MESURE, 0.95)
- Cherny (150 PRs/jour record), Curran top 5% (6× median = 18× baseline), Karpathy "peaks higher than 10×" —fait_partie_de→ Tail élite 10×+ (CONCEPT, 0.93)
- Writing muscle atrophy —converge_avec→ étude Anthropic junior engineers deskilling (CONCEPT, 0.91)
- Agents IA —permet→ adressage de la long-tail des besoins métier (CONCEPT, 0.94)
- Quality control traditionnel —est_basé_sur→ rythme humain (CONCEPT, 0.93)
- Alexandre Frizzo —affirme_que→ les coûts cognitifs sont cachés derrière la productivité affichée (AFFIRMATION, 0.94)

---
Canonical: https://www.thekb.eu/en/fiches/frizzo-linkedin-year-claude-code-output-doubled-attention-span-2026-05-05/