Imagine seeing a long-haired person with white hair and instantly thinking, “Rashōmon?” No deliberate joke-writing. No long search. The line simply appears—even when you are short on sleep.
That raises a surprisingly large set of questions. What is this instant-output ability? Can it be trained? Which abilities collapse under sleep loss and which sometimes survive? Can coffee or an energy drink restore them? Can sleep deprivation imitate pieces of attention-deficit/hyperactivity disorder (ADHD) or autism spectrum disorder (ASD)? And what should we make of extreme short-sleeper claims?
Researchers have, once again, been unreasonably specific.
1. Five-second answer: practical creativity has four moving parts, while sleep is the platform underneath
A useful breakdown is:
- Association — move quickly through semantic memory.
- Analogy — map the current situation onto a structurally similar case from another domain.
- Selection — keep the candidate that is both novel and relevant.
- Compression — reduce the whole chain to one line people can reconstruct themselves.
A 2024 study found that novelty depended more strongly on semantic associative ability, while appropriateness depended more strongly on executive control.[1] A 2026 Cognition study found that correct solutions to remote-association problems were often already present in participants’ first fast, intuitive responses.[2]
So an idea that “just appears” can still be genuine creative cognition.
Sleep, however, is a different layer. A burst of association can survive while vigilance, sequencing, or stable judgment deteriorate.
Even a creation god needs server maintenance.
2. Association game + analogy game + choosing the winner is basically the right model
Association is the ability to travel farther through the network of meanings. “White hair” might normally lead to “aging.” A broader search can move through “old woman,” “hair,” “classic literature,” and eventually “Rashōmon.” Semantic associative ability predicted idea novelty in the 2024 study.[1]
Analogy is different from mere word similarity. It asks: “What has the same structure in another domain?” A restaurant scene may map onto literature; a relationship problem onto quality control; a work process onto a game system.
Selection prevents broad association from becoming random noise. If fifty candidates appear, the useful skill is recognizing which distant candidate has multiple real points of contact and fits the conversation.
Compression turns the explanation into a usable line. “This visual configuration resembles a literary old-woman motif” is accurate but dead. “Rashōmon?” lets the listener reconstruct the rest.
Practical creativity is not simply producing distant ideas. It is producing, mapping, ranking, and compressing them fast enough to use.
3. Control is necessary, but an early checkpoint can stop the bulldozer before it moves
A 2025 study of 384 participants found that both free and controlled association contributed positively to originality, but stronger controlled association was linked to fewer ideas.[3]
Think of two departments:
- generation = bulldozer, opening possible routes;
- evaluation = surveyor, deciding which routes are safe and useful.
Trouble starts when the surveyor stands in front of the bulldozer and rejects every route before exploration begins.
This is exactly what rigid perfectionism can look like. Make a clay object, decide “wrong,” crush it. Make another, crush it. The prototype never survives long enough to become a product.
Research distinguishing realistic excellence from perfectionistic standards found that excellence-oriented participants produced more and more original divergent-thinking responses than perfection-oriented participants.[4]
But evaluation itself is not automatically toxic. A preregistered 2026 study with 440 participants found no significant originality or fluency differences between people who evaluated every idea immediately, evaluated after the task, or were not prompted to evaluate.[5]
The better conclusion is: monitoring is not the enemy; rigid rejection before exploration can be.
4. Can people train this? Yes, but it is not a cheat code
A 2024 meta-analysis covering 169 creativity-training studies and 844 effect sizes estimated an unadjusted effect of 0.53 standard deviations. After correcting for strong publication bias, estimates fell to about 0.29–0.32.[6]
So training can work, but the evidence does not support “everyone becomes a creative genius in a week.”
A 2025 university-student meta-analysis of 29 studies and 109 effects reported a medium overall effect, Hedges’ g=0.628.[7] Given the publication-bias problem in the broader literature, conservative expectations are appropriate.
For this specific instant-output skill, a research-informed practice set would be:
- Association: take one word and free-associate for 30 seconds without censoring.
- Analogy: ask what the day’s event resembles in literature, games, work, history, or another field.
- Selection: keep distant ideas only when at least two meaningful links exist.
- Compression: reduce the explanation to a short line after the candidate is selected.
That exact four-part routine has not itself been validated as a single randomized intervention; it is an application built from the component research.
5. Sleep loss first attacks the ability to stay reliably online
A 2024 meta-analysis of 44 studies on one night of restricted sleep opportunity—2 to 6 hours—found a large increase in sleepiness and clear impairment in sustained attention, including slower responses and more lapses.[8] Some other cognitive domains, including working memory and inhibitory control, did not show significant average impairment in that particular analysis.
The subjective experience is therefore not always “everything is 30% slower.” It can be more like:
95 → 20 → 90 → 40.
Performance becomes unstable.
With more severe or extended sleep loss, the damage broadens. A 2025 meta-analysis of 79 publications found impairments across working memory, inhibition, and cognitive flexibility.[9]
Chronic restriction is especially deceptive. In the classic 14-day experiment, 4- or 6-hour sleep opportunities produced cumulative performance deficits while subjective sleepiness failed to worsen at the same rate.[10]
Brain: getting worse.
Person: “I think I adapted.”
Those are not the same measurement.
6. And yet sleep-deprived “Rashōmon?” is plausible: a 2026 study landed very close to this question
In May 2026, Kyushu University researchers studied 27 young men. One group had only a 2-hour time-in-bed opportunity; the other slept normally.[11]
The sleep-restricted group became sleepier and showed worse sustained attention. Yet the groups did not differ clearly on a verbal divergent-association task or tangram insight problems. Bayesian analyses supported the possibility that these aspects of creativity were preserved under that acute partial deprivation.[11]
So “I’m exhausted but the weird association still arrives instantly” is plausible.
What it does not show is that two hours of sleep is safe or that all cognition remains intact. The paper itself warns that other cognitive functions and health can still be compromised.[11]
The creative engine may still be running while the lane-keeping system is failing.
7. Caffeine power-up: what comes back, and what often does not
A meta-analysis of 45 publications and 327 effect estimates found that caffeine after sleep loss improved attention response speed and accuracy, information processing, driving performance, and—more modestly—executive function.[12]
Caffeine is therefore good at restoring wakefulness-like performance: staying alert, reacting faster, and reducing lapses.
It is not the same as restoring the entire sleep-deprived brain.
In a 276-person experiment, 200 mg caffeine after an all-nighter improved psychomotor vigilance but did not significantly restore a complex placekeeping task for most participants.[13]
Caffeine: Awake! Faster reactions!
Brain: By the way, we skipped step three.
Creativity shows another split. In a double-blind trial of 88 people, 200 mg caffeine improved convergent problem solving—the kind with one correct solution—but did not significantly improve divergent idea generation or working memory.[14]
So coffee has no proven “Super Rashōmon Mode.”
A better summary is:
association generator: not clearly boosted;
attention that catches and uses the generated idea: easier to boost.
8. Red Bull versus coffee: much of the wing is caffeine
As of September 2026, Red Bull Japan states that a 250 mL can contains 80 mg caffeine.[15]
The current Japanese ingredient page lists caffeine, arginine, B vitamins, sugars, and water among the main ingredients. It is therefore a mistake to automatically import findings from overseas taurine-containing formulations into the current standard Japanese product.[15]
For cognition, the clearest common active ingredient between coffee and Red Bull is caffeine.
But the buff can attack the next night’s maintenance window. A 24-study meta-analysis found that caffeine reduced total sleep by about 45 minutes, reduced sleep efficiency by 7%, delayed sleep onset by 9 minutes, increased wake after sleep onset by 12 minutes, and reduced deep sleep.[16]
Its model estimated that a 107 mg/250 mL coffee would need to be consumed about 8.8 hours before bedtime to avoid reducing total sleep time on average.[16] Individual sensitivity varies.
Sleep loss → caffeine → shorter next sleep → more caffeine is a perfectly buildable negative loop.
You gained wings and deleted the landing schedule.
9. Maybe everyone online saying ADHD or autism is just sleep-deprived? Partly plausible as symptom noise, wrong as a diagnosis claim
The joke points to a real diagnostic problem.
CDC guidance states that ADHD cannot be diagnosed by one test and that sleep disorders, anxiety, depression, and learning problems can produce similar symptoms.[17] A review of sleep-disordered breathing likewise notes that sleep problems can worsen existing ADHD symptoms and produce cognitive deficits that mimic ADHD symptoms.[18]
Sleep loss can produce pieces such as poor concentration, forgetfulness, more mistakes, slowed responses, and emotional instability. A checklist does not identify the cause.
So:
five hours of sleep → poor concentration → online checklist → “ADHD!”
is not a causal diagnosis.
But ADHD is not “just sleep deprivation.” Clinical diagnosis considers persistence, impairment, developmental course, and symptoms across settings.
Autism is even less reducible to a bad week of sleep. CDC’s DSM-5 summary requires persistent social-communication differences plus restricted/repetitive patterns, with symptoms present in the early developmental period.[19]
Sleep deprivation can worsen social functioning or make fragments look similar. It does not by itself establish a neurodevelopmental disorder.
A symptom list is a dashboard warning light, not a mechanic’s final report.
10. What about 30-minute short sleepers? Separate public claims from natural-short-sleeper science
Japanese public figure Daisuke Hori has publicly described sleep schedules in the range of roughly 30–45 minutes. In a published interview he described originally needing about eight hours, training himself to shorten sleep, and at that time using about 45 minutes of main sleep plus two 15-minute power naps.[20]
A public claim is not the same thing as an independently verified physiological sleep requirement.
Natural short sleepers do exist. Familial natural short sleep has been associated with variants involving DEC2/BHLHE41, ADRB1, NPSR1, and GRM1.[21] But the well-described phenotype is typically around 4–6 hours, not 30 minutes.[21]
In the public evidence checked for this article, no peer-reviewed long-term study was found in which Hori’s total 24-hour sleep was continuously verified with EEG and then compared with normal sleepers across cognition, metabolism, and cardiovascular outcomes.
The appropriate status is therefore:
an extreme public claim of long-term ultra-short sleep,
not
scientific proof that ordinary people can train themselves to need 30 minutes.
This distinction matters because chronic sleep restriction can produce accumulating objective deficits while people feel they have adapted.[10]
Japan’s Ministry of Health, Labour and Welfare currently recommends that adults secure the sleep they need, using at least six hours as a general target, while acknowledging individual differences.[22]
If somebody truly maintains every relevant health and cognitive outcome on 30 minutes for years, that is not a reason to skip science. It is a reason to bring in the sleep lab.
Sleep scientist: “Great. Now wear the EEG.”
11. Conclusion: sleep is maintenance, caffeine is a temporary buff, and creativity has partially separable modules
The whole conversation fits one model.
Creativity combines association, analogy, selection, and compression.[1][2]
Control is useful, but rigid screening before exploration can narrow generation; meanwhile, evaluation itself is not automatically harmful.[3][5]
Sleep loss hits vigilance and stability early, then spreads into executive functions as severity and duration increase.[8][9][10]
Some divergent thinking and insight can survive acute partial sleep loss under specific experimental conditions.[11] That makes “sleepy but still instantly funny” plausible.
Caffeine can strongly restore alertness and reaction performance but does not fully restore procedural control and has little evidence for boosting divergent creativity.[12][13][14]
ADHD/autism self-diagnosis must account for sleep as a source of overlapping symptoms without equating sleep deprivation with neurodevelopmental disorders.[17][19]
Natural short sleepers are real, but their research is not proof that 30-minute sleep is trainable for the general population.[21][22]
So the final model is simple:
Sleep = system maintenance.
Coffee = temporary stat buff.
Creativity = a module that can sometimes keep firing while other systems are already dropping packets.
Getting “Rashōmon?” instantly while sleep-deprived is not evidence that the rest of the brain is fine.
Even the creation god sleeps.
Sources
[1]–[22] map to the language-neutral Global References at the end of this file.
Sources
- Wang X, et al. “Semantic associative abilities and executive control functions predict novelty and appropriateness of idea generation.” Communications Biology. 2024;7:703. PMID 38849461. DOI: — https://pubmed.ncbi.nlm.nih.gov/38849461/ doi.org
- Beucler J, De Neys W. “Intuitive insight: Fast associative processes drive sound creative thinking.” Cognition. 2026;271:106422. PMID 41500017. DOI: — https://pubmed.ncbi.nlm.nih.gov/41500017/ doi.org
- Creative spark or overthinking? The synergistic effects of free and controlled associations on creativity.” Thinking Skills and Creativity. 2025;57:101818. DOI doi.org
- Gaudreau P, et al. “Is perfectionism a killer of creative thinking? A test of the model of excellencism and perfectionism.” British Journal of Psychology. 2022. PMID 34472100. DOI: — https://pubmed.ncbi.nlm.nih.gov/34472100/ doi.org
- Does reflecting one's own ideas affect creativity? Metacognitive monitoring activation shows no effect on divergent thinking.” Learning and Individual Differences. 2026;129:102872. DOI doi.org
- Sio UN, Lortie-Forgues H. “The impact of creativity training on creative performance: A meta-analytic review and critical evaluation of 5 decades of creativity training studies.” Psychological Bulletin. 2024;150(5):554–585. PMID 38635185. DOI: — https://pubmed.ncbi.nlm.nih.gov/38635185/ doi.org
- Guo H, Zhou Z, Ma F. “Effect of training programs on the creativity of university students: A meta-analysis.” Thinking Skills and Creativity. 2025;56:101779. DOI doi.org
- Impact of one night of sleep restriction on sleepiness and cognitive function: A systematic review and meta-analysis.” Sleep Medicine Reviews. 2024. PMID 38759474. DOI: — https://pubmed.ncbi.nlm.nih.gov/38759474/ doi.org
- Cao Y, et al. “The impairments of sleep loss on core executive functions: A systematic review and meta-analysis.” Sleep Medicine Reviews. 2025. PMID 40946426. DOI: — https://pubmed.ncbi.nlm.nih.gov/40946426/ doi.org
- Van Dongen HPA, et al. “The cumulative cost of additional wakefulness: Dose-response effects on neurobehavioral functions and sleep physiology from chronic sleep restriction and total sleep deprivation.” Sleep. 2003;26(2):117–126. PMID 12683469. DOI: — https://pubmed.ncbi.nlm.nih.gov/12683469/ doi.org
- Motomura Y, et al. “Partial sleep deprivation with a 2-h time-in-bed may not change divergent thinking and insight problem solving in young males.” Journal of Physiological Anthropology. 2026;45:18. PMID 42169155. DOI: — https://pubmed.ncbi.nlm.nih.gov/42169155/ doi.org
- Systematic review/meta-analysis of acute caffeine consumption following sleep loss; 45 publications / 327 effect estimates. PMID 31837359 pubmed.ncbi.nlm.nih.gov
- Stepan ME, Altmann EM, Fenn KM. “Caffeine selectively mitigates cognitive deficits caused by sleep deprivation.” Journal of Experimental Psychology: Learning, Memory, and Cognition. 2021;47(9):1371–1382. PMID 34014758. DOI: — https://pubmed.ncbi.nlm.nih.gov/34014758/ doi.org
- Zabelina DL, Silvia PJ. “Percolating ideas: The effects of caffeine on creative thinking and problem solving.” Consciousness and Cognition. 2020;79:102899. PMID 32086187. DOI: — https://pubmed.ncbi.nlm.nih.gov/32086187/ doi.org
- Red Bull Japan. Current Japanese product ingredient/caffeine information. 250 mL standard can: 80 mg caffeine. and https://www.redbull.com/jp-ja/energydrink/questions/how-much-caffeine-is-in-a-can-of-red-bull-energy-drink redbull.com
- Gardiner C, et al. “The effect of caffeine on subsequent sleep: A systematic review and meta-analysis.” Sleep Medicine Reviews. 2023;69:101764. PMID 36870101. DOI: — https://pubmed.ncbi.nlm.nih.gov/36870101/ doi.org
- U.S. Centers for Disease Control and Prevention. “Diagnosing ADHD. cdc.gov
- Ivanov I, et al. Review on sleep-disordered breathing and ADHD risk/symptom overlap. Journal of Attention Disorders. 2024;28(5):686–698. PMID 38353411 pubmed.ncbi.nlm.nih.gov
- U.S. Centers for Disease Control and Prevention. “Clinical Testing and Diagnosis for Autism Spectrum Disorder. cdc.gov
- PERSOL. Public interview with Daisuke Hori, 2024: “1日45分睡眠のショートスリーパー堀大輔さんが語る、人生を豊かにする短眠術. studio.persol-group.co.jp
- Reviews of familial natural short sleep and sleep genetics, including DEC2/BHLHE41, ADRB1, NPSR1, and GRM1-related findings. and https://pubmed.ncbi.nlm.nih.gov/34722012/ pmc.ncbi.nlm.nih.gov
- 厚生労働省. 「健康づくりのための睡眠ガイド2023」 mhlw.go.jp
