ISO on the Outside, Absolute Nonsense on the Inside — What “Dopagaki Driven Development” and “Pachinko–Dopamine–Charge–Atonement” Teach Us About Incongruity, Manzai, Museum Labels, and AI Humor

Telling an AI to “make it funny” often produces laughter cues instead of humor: slang, exclamation marks, and the textual equivalent of someone laughing at thei

Advertisement
Advertisement

TL;DR: Raise the formality, lower the subject matter, keep the logic weirdly excellent

Telling an AI to “make it funny” often produces laughter cues instead of humor: slang, exclamation marks, and the textual equivalent of someone laughing at their own joke. A more reliable mechanism is incongruity: build semantic world A, drop the reader into distant world B, but preserve one logical bridge that makes the jump intelligible after the fact.[1][2] Register humor gives an especially useful version: most of a text stays in one style while a small element comes from a radically different register.[3]

Our two specimens are DDD: Dopagaki Driven Development [4], a parody of Domain-Driven Design, and PDCA: Pachinko / Dopamine / Charge / Atonement, rendered in Japanese as “pachinko, dopamine, spending, regret.” The trick is not to vandalize professional terminology. It is to preserve the serious shell with unreasonable care while replacing the contents with something painfully mundane.

The format is ISO-grade. The contents have failed every life-decision audit. Somehow, QA passes.

1. A PDCA cycle that improves nothing

The real PDCA is Plan–Do–Check–Act, a four-step cycle repeated for continuous improvement.[5] The parody preserves not only the initials but also the loop:

Pachinko → dopamine → spending → regret → pachinko again.

  • Classification: cyclical decision model
  • Habitat: station districts and entertainment streets
  • Reproductive phrase: “Today I can win it back.”
  • Predators: bank notifications, morning, household budgets
  • Conservation status: still operational

It functions as a PDCA cycle. It simply refuses to improve anything. A random P-D-C-A word list only matches letters; this version preserves causality and recursion. It is also close to a backronym, where an existing name is later expanded into matching initials.[6]

2. DDD, now with dramatically worse governance

Real Domain-Driven Design is a software design approach centered on the business domain, with concepts such as strategic/tactical design, bounded contexts, domain models, and ubiquitous language.[7]

Then comes Dopagaki Driven Development. The Japanese slang is preserved because the register itself matters. The method is presented with a three-letter acronym, English expansion, engineering-methodology tone, and technical slide. Its idea is essentially: trust all AI-generated code (?) and chase maximum speed.

The tiny “(?)” is a second joke. Without it, the method is merely reckless. With it, reason returns for half a second, asks whether this might be a terrible idea, and then leaves without changing the policy.

  1. Domain-Driven Design is expected; Dopagaki Driven Development appears.
  2. The methodology seems reckless.
  3. The author appears to notice.
  4. The acceleration continues.

The professional form is not broken. It is over-respected.

3. Incongruity needs a bridge back

Incongruity theories focus on violations of expectation.[1] But randomness is not enough.

“The meeting started. The refrigerator became Jupiter” is unexpected but has no bridge.

“We introduced AI and cut task time in half. We used the saved time for a two-hour meeting about how to use AI” works because:

  • A: AI → efficiency.
  • B: AI → implementation meetings.
  • Bridge: technology really can create training, coordination, and governance overhead.

In Suls-style incongruity-resolution accounts, the audience detects the mismatch and then finds a reinterpretation that makes the punch line fit the earlier material.[1]

“That makes no sense” → half a second → “Unfortunately, it does.”

The General Theory of Verbal Humor likewise considers script opposition together with a logical mechanism, situation, target, narrative strategy, and language.[2][8]

4. Formality versus triviality: register humor

Venour, Ritchie, and Mellish describe register humor as text that mostly belongs to one style or tone while a small part is radically different in register.[3]

This specimen exhibits increased activity after dusk and has been observed acquiring pudding from convenience stores.

The event is trivial; the prose belongs in a natural-history museum.

That same distance powers:

The Demon King pays neighborhood association dues.
God joins the condominium management board.
The world is managed in Excel.
Death is exported to PDF.
The era is sent to committee review.

Formality creates altitude. Mundanity creates the drop. Logic ensures a landing.

5. Through the lens of manzai

Japanese scholarship on manzai has analyzed laughter in terms including errors, inefficiency, and inconsistency.[9] A boke-like move deviates from the normal model; a tsukkomi-like move repairs it; the next line breaks the repair again.

A: “AI will cut our workload in half.”
B: “Great.”
A: “So we’ll hold three AI adoption meetings a week.”
B: “You increased the workload.”
A: “AI writes the minutes, though.”
B: “Then cancel the meeting.”
A: “We need a meeting to decide that.”
B: “Just give the company to the AI.”

In prose, the reader can supply the tsukkomi internally:

A new committee was established to consider reducing the number of committees.

Explaining the joke too aggressively fires the reader from the tsukkomi department.

6. Structural preservation is the secret

DDD preserves the three-letter acronym, “Driven Development” naming pattern, methodology pose, and authority of a technical presentation. PDCA preserves four ordered stages, causal progression, recursion, and repeatability.

Criterion Weak Strong
Semantic distance adjacent domains distant domains
Structure words only order, causality, procedure
Resolvability random makes sense afterward
Explanation explains punch line leaves room
Tone laughs at itself stays deadpan

Corrupt one premise. Keep the subsequent reasoning correct.

7. Why museum labels are comedy machines

Museum prose already offers rigid categories: name, classification, range, habitat, behavior, diet, reproduction, conservation status.

Meeting (Meetingus endlessensis)

A social ritual widely distributed across modern organizations. Four to twelve individuals gather physically or online to exchange information. Reduced participant count does not reliably reduce duration. When no conclusion is reached, the group may reproduce through the phrase “Let’s schedule a separate meeting.”

  • Classification: repetitive group ritual
  • Habitat: conference rooms, Teams, Zoom
  • Predators: clear decision rights, agendas, hard-stop times
  • Conservation status: no evidence of extinction

The museum register explains everything except why it is funny. That is the point.

8. Literature, umbrellas, and one surviving bridge

A normal umbrella review says it blocks sun and rain. A shifted frame says:

Humanity failed to control the weather, so it settled for controlling a circle roughly one meter wide.

Civilization and umbrellas are distant domains, but the bridge is real: an umbrella creates a tiny local protective zone.

Treat literature about enduring rain, wind, or summer heat as a human requirements specification, then ask which requirements a modern umbrella can implement in hardware. Now literature × requirements engineering × product review collide.

  • rain: potentially compliant
  • summer heat: depends on shading/thermal performance
  • freedom from desire: not addressable by product hardware
  • always smiling quietly: no facial subsystem implemented

The joke is not contempt for literature. The quality-assurance department is simply taking the specification with terrifying sincerity.

9. Benign violation: increase distance, not cruelty

McGraw and Warren’s Benign Violation Theory proposes that humor can arise when something is perceived simultaneously as a violation and as benign or acceptable.[10] “The Demon King pays neighborhood dues” violates expectations but causes no serious harm. Move too close to real injury, humiliation, or personal attack and discomfort can overwhelm the incongruity.

Increase semantic distance, not cruelty.

A 2021 integrative review notes more than twenty psychological theories of humor and finds that surprise alone is not enough to distinguish humorous from non-humorous experience.[11] Thus “formal register × trivial content × preserved logic” is a useful engineering pattern, not a universal law of comedy.

10. How to delegate this to AI

Keep all facts unchanged. Do not manufacture humor with “lol,” excessive slang, or laugh markers. Establish the semantic domain the reader naturally expects, then connect it to a distant domain so the mismatch initially feels wrong but becomes logically interpretable. Try high-register frames such as museum labels, bureaucracy, quality management, requirements specifications, academic papers, law, and executive reports. Corrupt only one premise; keep subsequent reasoning, causality, and tone serious. Generate 20 candidates and score them on semantic distance, structural preservation, processing speed, and safety. Keep only the top three. Do not explain the punch line.

Workflow: lock facts → define world A → generate twenty remote worlds B → find one legitimate bridge → preserve source structure → maintain deadpan → delete explanations → QC facts, cruelty, and repetition.

AI is often more useful as a semantic-space search engine that proposes twenty strange bridges than as a comedian expected to hit one perfect punch line.

11. Failure modes

  • No bridge: “An umbrella opened and the stock market became Mars.”
  • Too close: “The meeting was long and tiring.”
  • Initials only: four unrelated PDCA words.
  • Explaining the joke: “Conservation status: still operational (lol).” Specimen deceased.
  • Fake formality: random jargon does not create academic register.

12. Reusable templates

  • Mythic being × household institution: The Demon King pays neighborhood dues.
  • Cosmic concept × office operation: The world is managed in Excel; death is exported to PDF.
  • Trivial choice × quality system: Run an FMEA before choosing lunch.
  • Daily object × requirements engineering: Define an umbrella as a local meteorological-resistance device.
  • Ordinary failure × bureaucracy: “Overslept” → “An operational issue concerning optimization of wake-time control was identified.”
  • Anything × museum: classification, habitat, reproduction, predators, conservation status.

13. Conclusion: corrupt the premise, not the reasoning

The strong combination is formality × mundanity × structural preservation × resolvability × deadpan delivery. DDD works because the shell of Domain-Driven Design survives its mutation into Dopagaki Driven Development. PDCA works because the loop itself survives the mutation into Pachinko–Dopamine–Charge–Atonement.

For AI writing:

Search for distant semantic domains, preserve one logical bridge, keep the source structure, and forbid the narrator from laughing first.

  • Classification: structure-preserving incongruity humor
  • Habitat: technical slides, articles, manzai, social media, corporate decks
  • Diagnostic feature: one abnormal premise, otherwise normal reasoning
  • Conservation status: expanding
  • Quality verdict: ISO-grade form. Catastrophic contents. Approved.

Sources

  1. Suls-style incongruity-resolution overview and later discussion. Incongruity-resolution cases in jokes, Lingua sciencedirect.com
  2. Salvatore Attardo. The General Theory of Verbal Humor. In The Linguistics of Humor: An Introduction, Oxford University Press, 2020 doi.org
  3. Chris Venour, Graeme D. Ritchie, Chris Mellish. Dimensions of incongruity in register humour. In The Pragmatics of Humour across Discourse Domains, 2011, pp. 125–144 doi.org
  4. Public social-media example that prompted the DDD discussion, posted 2026-08-30 x.com
  5. American Society for Quality (ASQ). What is the Plan-Do-Check-Act (PDCA) Cycle? asq.org
  6. Merriam-Webster. backronym merriam-webster.com
  7. Microsoft Learn, Azure Architecture Center. Use Domain Analysis to Model Microservices — overview of DDD, strategic/tactical design, bounded contexts, and ubiquitous language learn.microsoft.com
  8. Salvatore Attardo, Victor Raskin. Script theory revis(it)ed: joke similarity and joke representation model. HUMOR, 4(3–4), 1991, 293–348 doi.org
  9. 中田一志. 漫才の笑い:エラーと非効率性と非整合性. 『日本語・日本文化』41, 大阪大学日本語日本文化教育センター, 2014, pp. 1–35 doi.org
  10. A. Peter McGraw, Caleb Warren. Benign Violations: Making Immoral Behavior Funny. Psychological Science, 21(8), 2010, 1141–1149 doi.org
  11. Caleb Warren, Adam Barsky, A. Peter McGraw. What Makes Things Funny? An Integrative Review of the Antecedents of Laughter and Amusement. Personality and Social Psychology Review, 25(1), 2021, 41–65 doi.org
Advertisement
Mendoi-chan

Written by

Mendoi-chan

She turns friction at work and in everyday life into clear structure and practical next steps.

About
Advertisement

Latest articles

  1. 1Can Rules Improve AI Writing? — Turning a Professional Editor into a TypeScript Workflow
  2. 2What Makes a Person “Deep”? — They Resist Easy Answers, Keep Complexity Intact, and Still Converge on a Judgment
  3. 3References / 参考文献
  4. 4English|Why does my car smell burnt even before I start the engine? How to separate trapped smoke from a real fault
  5. 5English (en)

You may also like

Advertisement