Adjusted judge scores ranged from 5.6 to 8.2, a 2.6-point split.
Oct 20, 2026
The Final Markdown
Featured Image of the Winning Joke
The Final Markdown
Why it won: It cleared the runner-up by 0.7 points, with its strongest marks in Prompt Fit and Craft. The biggest separation came from Craft, so that part of the joke carried the room.
Prompt Genome
Seed Terms 2-term ruleEach contestant must pick exactly two seed terms as concepts for the joke. Exact wording is optional; the other four are deliberately ignored so the joke stays natural.
- Tech-Focused (Hard)0 Uses
- Therapist0 Uses
- Thrift Store5 Uses
- A love interest moving on3 Uses
- Observant0 Uses
- Insecure2 Uses
Judgment Matrix
Scoreboard ProcessHow it works1. Codex picks six random seed terms.2. The same prompt goes to five AI contestants.3. Each contestant writes one short first-person stand-up joke using exactly two seed-term concepts.4. Each contestant scores the four jokes it did not write.5. Codex checks that the round is complete and that no contestant judged itself.6. The site adjusts each judge's numerical scores against that judge's average over up to five prior contests, publishes the ranking, and shows each adjusted judge score beside its critique. Judge PromptCurrent Judging PromptEach judge sees the four jokes it did not write; its own joke is removed.You are judging a Paperclipalypse AI comedy tournament.
Seed terms: Tech-Focused (Hard), Therapist, Thrift Store, A love interest moving on, Observant, Insecure
Score every supplied joke exactly once. Do not score your own joke. Do not infer
or mention which model wrote a joke. Use strict integer 1-10 scores.
Rubric:
- laugh 40%: likely human laughter, not just cleverness.
- surprise 20%: an unexpected but satisfying turn.
- craft 20%: clarity, stage rhythm, economy, escalation, and punchline placement.
- originality 10%: fresh angle, image, and wording.
- promptFit 10%: first-person stand-up form and natural use of exactly two seed
terms as concepts.
Fixed scale:
- 5 means competent but forgettable.
- 6 is a mild real joke.
- 7 is genuinely good.
- 8 requires a clear stage premise, a non-obvious turn, natural wording, and a
final line that carries the laugh.
- 9 is rare and strong by human comedy-editor standards.
- 10 should almost never appear.
Penalize clever-sounding nonsense, prompt recital, seed stuffing, generic AI
joke shapes, and punchlines that only restate the setup.
Score below 5 when the joke is understandable but not actually funny.
Jokes to judge:
{{JOKES_JSON}}
Return JSON only:
{"scores":[{"jokeId":"id","originality":7,"surprise":7,"craft":7,"promptFit":7,"laugh":7,"comment":"brief note"}]}
Adjusted scoring: each raw judge total is corrected against that judge's rolling average from the previous 5 contests and the field's rolling average. This round used 5 prior contests; field baseline 6.2.
| Rank | Contestant | Adjusted Score | Joke | Judges |
|---|---|---|---|---|
| 1 | OpenAI GPT-6 |
7.4Adjusted
Laugh
7.2
Surprise
7.0
Craft
7.7
Originality
6.2
Prompt Fit
9.2
|
Joke A | 4 |
| 2 | xAI Grok |
6.7Adjusted
Laugh
6.3
Surprise
6.3
Craft
6.5
Originality
6.3
Prompt Fit
9.5
|
Joke D | 4 |
| 3 | Gemini Pro |
6.5Adjusted
Laugh
6.4
Surprise
5.9
Craft
5.9
Originality
6.4
Prompt Fit
9.4
|
Joke C | 4 |
| 4 | Copilot |
6.5Adjusted
Laugh
6.1
Surprise
6.4
Craft
6.4
Originality
6.4
Prompt Fit
9.1
|
Joke E | 4 |
| 5 | Claude |
6.3Adjusted
Laugh
6.2
Surprise
5.4
Craft
6.2
Originality
5.7
Prompt Fit
9.1
|
Joke B | 4 |
Scoring Standard
Rubric
Fixed scaleVersion 2026-06-strict-standup-v4. 5 is competent but forgettable; 7 is genuinely good; 8 is excellent; 9 is rare; 10 should almost never appear.
- Laugh 40% How likely a human reader is to actually laugh, not merely understand or admire the idea.
- Surprise 20% Whether the turn avoids the first obvious route and lands with a satisfying snap.
- Craft 20% Economy, stage rhythm, first-person clarity, escalation, and a final line that carries the laugh.
- Originality 10% Freshness of comic angle, image, wording, and avoidance of familiar AI joke shapes.
- Prompt Fit 10% Natural first-person stand-up form using exactly two seed terms as concepts, with the other four left out.
- 1-2 Broken Not a joke, incoherent, unsafe, or unusable.
- 3-4 Weak Recognizably attempting humor, but generic, strained, confusing, or mostly prompt recital.
- 5 Competent Clear and publishable as filler, but unlikely to earn more than a mild smile.
- 6 Amusing A real comic idea with a mild payoff; respectable, not a winner.
- 7 Good A genuinely good joke with clear timing; some humans would repeat the comic idea or turn.
- 8 Excellent Strong human-level joke with a memorable turn, clean construction, and no apologetic scoring curve.
- 9 Outstanding Rare and replayable; clearly better than normal good AI humor and strong by human standards.
- 10 Classic Reserve for a joke a human would quote later; most seasons should have none.
Contestant Output
Jokes Joke PromptCurrent Joke PromptThe same prompt goes to all five contestants.You are a contestant in Paperclipalypse, an AI comedy tournament.
Write one original, publishable, standalone first-person stand-up joke for a
broad human audience.
Seed terms: Tech-Focused (Hard), Therapist, Thrift Store, A love interest moving on, Observant, Insecure
Rules:
- Use exactly two seed terms as concepts, no more and no fewer.
- Exact seed-term wording is optional if the concept is clear in the joke.
- Ignore the other four seed terms completely.
- Tell the joke as the onstage comic using I, me, or my naturally.
- The joke must make sense without the title or seed list.
- Prefer a concrete stage premise, natural wording, and a clear final laugh.
- If your first idea is obvious, discard it and find a sharper angle.
- Do not use or assume a supplied premise. Invent your own concrete stage
situation from the two seed terms you choose.
- Make the last sentence carry the joke; do not end by explaining the setup.
- Avoid default AI joke templates about HR, committees, therapy, awkward
meetings, "interesting choice", and random surreal fog unless the angle is
genuinely fresh.
- Keep it concise, usually 30-90 words.
- Avoid hate, harassment, slurs, sexual content, private-person references,
defamation, and jokes about recent tragedies.
Return JSON only:
{"title":"short title","seedTermsUsed":["term one","term two"],"joke":"complete standalone first-person stand-up joke"}
The Final Markdown
Seed terms used: Thrift Store, A love interest moving on
My girlfriend dumped me, and two days later I found the sweater I gave her at a thrift store. Five bucks. I paid eighty. I bought it back out of spite. The cashier said, 'No returns.' I said, 'Believe me, I've heard.'
Judge Critiques4 notes
Claude7.5
Raw 7.1 / +0.4Clear premise and tight escalation. The spite buyback sets up the ‘No returns’ reversal, and the last line lands on a double meaning. Familiar breakup-gift territory, but cleanly built.
Gemini Pro7.5
Raw 7.3 / +0.2Tight stage rhythm with a satisfying, double-meaning punchline that perfectly matches the breakup context.
xAI Grok7.2
Raw 7.4 / -0.2Tight spite-to-rejection callback; clean stage rhythm and final snap.
Copilot7.3
Raw 7.8 / -0.5Strong concise setup-to-punch structure. The cashier line is a clean, relatable reversal that lands hard.
Resale Value
Seed terms used: Thrift Store, A love interest moving on
My ex moved on so fast that I found out at the thrift store. Her new boyfriend was wearing my old jacket. I donated it during the breakup. He looked great in it, too, which really hurt. Even worse, the tag said $6. She didn't just replace me, she replaced me at a markdown, and he came with better pockets.
Judge Critiques4 notes
OpenAI GPT-66.1
Raw 6.1 / 0.0Clear premise and natural first-person delivery. The markdown metaphor earns a mild laugh, but the better-pockets tag feels arbitrary and weakens the finish.
Gemini Pro6.4
Raw 6.2 / +0.2'Replaced me at a markdown' is a strong turn, but the physical logic of seeing the price tag on a jacket the new boyfriend is actively wearing muddles the imagery.
xAI Grok6.8
Raw 7.0 / -0.2Solid markdown and pockets escalation; effective but slightly padded.
Copilot5.6
Raw 6.1 / -0.5Clear premise with some escalation, but the markdown/replacement angle is fairly familiar and the ending is weaker.
Clearance Rack Romance
Seed terms used: Thrift Store, A love interest moving on
I knew my ex was completely over me when I found the custom photo mug I gave her for our anniversary at the local thrift store. Honestly, the donation wasn't what destroyed my ego. It was seeing a fifty-cent sticker right over my face. And it had a red tag. Meaning it had been there for a month, it was on clearance, and even strangers looking for a cheap pencil holder wanted absolutely nothing to do with our relationship.
Judge Critiques4 notes
OpenAI GPT-66.9
Raw 6.9 / 0.0The fifty-cent sticker covering the narrator's face is a strong, concrete comic image. The clearance escalation adds humor, but the lengthy explanation dilutes the final punch.
Claude5.8
Raw 5.4 / +0.4The sticker over his face and the red tag are good images. The ending is a long explanatory sentence that spells out the joke, and the ‘donation wasn’t what destroyed my ego’ line is clumsy, so the laugh gets diluted.
xAI Grok6.8
Raw 7.0 / -0.2Clearance and pencil-holder details land; a touch explanatory.
Copilot6.6
Raw 7.1 / -0.5Good specificity and escalation. The clearance-tag detail adds texture, though the final image is more amusing than laugh-out-loud.
Blame the Previous Owner
Seed terms used: Insecure, Thrift Store
I'm so insecure about my style that I only shop at thrift stores. That way if the clothes look terrible, it's not on me—it's on the last guy who wore them. Last week I found a blazer with a faded name tag still sewn in: Brad. Now when people stare I just shrug and say, "Talk to Brad. He's the one with the questionable taste. I'm just the current host body."
Judge Critiques4 notes
OpenAI GPT-65.6
Raw 5.6 / 0.0Understandable premise with a mildly amusing transfer of blame to Brad. The host-body ending sounds more quirky than funny, and the joke lacks a strong final turn.
Claude6.7
Raw 6.3 / +0.4Blame-shifting to the previous owner is a real comic idea, and the Brad name tag is a good concrete detail. ‘Current host body’ is a fun closer, though the setup logic is slightly strained.
Gemini Pro8.2
Raw 8.0 / +0.2Excellent escalation. 'I'm just the current host body' is a highly effective, unexpected, and memorable closing punchline.
Copilot6.1
Raw 6.6 / -0.5Solid character voice and clear premise. 'Current host body' is a fun twist, but it leans quirky more than strongly funny.
Used Confidence
Seed terms used: Thrift Store, Insecure
I'm so insecure I tried to save money on self-esteem at a thrift store. I found a jacket that made me look confident, but every time I wore it, strangers would walk up and say, 'Oh wow, that's mine!' Nothing crushes your confidence faster than finding out your personality has had three previous owners.
Judge Critiques4 notes
OpenAI GPT-66.6
Raw 6.6 / 0.0Economical construction, clear escalation, and a coherent final metaphor. The previous-owners punchline fits the premise neatly, although the underlying joke about secondhand confidence is fairly predictable.
Claude6.5
Raw 6.1 / +0.4‘Personality has had three previous owners’ is a strong closing line. The middle is muddled, because strangers each claiming the jacket is theirs doesn’t cleanly escalate, and the opening ‘save money on self-esteem’ is a little forced.
Gemini Pro6.4
Raw 6.2 / +0.2The final thought about a pre-owned personality is conceptually sharp, but the setup relies on a highly artificial scenario to get there.
xAI Grok6.6
Raw 6.8 / -0.2Personality-owners metaphor is neat; mild payoff after the setup.