Escape Meridian

TechCrunch · Jul 12, 2024

TechCrunch reports individual writing gains alongside risk to collective novelty

Markdown · JSON · Original source

TechCrunch published this report on July 12, 2024. It describes a study by Anil Doshi and Oliver Hauser in which people writing short stories could use GPT-4 for story ideas. TechCrunch reports that participants with lower creativity scores received higher ratings when they used AI-generated ideas, while those with higher scores saw little to no benefit, or possibly worse ratings. The study also found that stories produced with AI were more similar to others in their category. Its authors warn that if publishers widely adopt AI-inspired stories, collective novelty may decline further, even as individual writers receive higher evaluations. The concern is a tension between a benefit for some writers and less varied output across a group—not that the stories were all alike. The study tested short stories in a controlled setting, and Hauser says real-world use will look different. The source does not establish how these results would translate to other creative work or publishing in practice.

Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.

TechCrunchArticle · techcrunch.comOpen original source

Original source

TechCrunch · Article

Source details

Original source date: Jul 12, 2024 (2024-07-12).

Date basis: Date from the publisher’s archive: its article address or archive listing.

Added to Escape Meridian: 2026-11-18T11:53:19.689Z. First observed: 2026-10-24T06:03:53.496Z. Retrieved: 2026-10-24T06:05:47.948Z.

Referenced by

Recorded citations in public readings. A reference alone does not establish agreement or influence.

No public references recorded here yet.

Reference data (JSON)

Limits

The study focused on short stories and used a controlled setting; Hauser says real-world implementation will look very different. Text as retrieved on October 24, 2026; originally published July 12, 2024.

Open question

How would the balance between individual writing gains and less varied stories change when generative AI is used in real-world creative work?

Source-stated

TechCrunch reports that the study found higher story ratings for participants with lower creativity scores when they used AI-generated ideas.

Given one AI-generated idea, they scored higher on every metric. Given the choice of five, they scored even higher.

Source-stated

TechCrunch reports that participants with higher creativity scores saw little to no benefit, or possibly worse ratings, with AI help.

those participants saw little to no benefit at all, or even (though it’s very close and arguably not significant) worse ratings.

Source-stated

The study found that stories produced with generative AI were more similar to others in their category, with a reported difference in the 9% to 10% range.

They found that access to generative AI caused the resulting stories to be closer to the average for their category. In other words, they were more similar

Source-stated

The study’s authors warn that wider adoption could reduce the collective novelty of stories, despite individual gains.

While these results point to an increase in individual creativity, there is risk of losing collective novelty.

Source-stated

Hauser says that real-world implementation will differ from the study’s controlled setting.

implementation of generative AI ‘in the wild’ will look very different than our controlled setting

Source and contributor context

TechCrunch · Publisher

The publisher is identified here. No individual author has been inferred.

Source media

Media retrieval status: unchecked.

Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.

Editorial provenance

Version 1. Publication ID: 099308c1-3545-d757-d7b2-3ec45a3fbeaf.

Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.

Source check: passed by anthropic/claude-opus-5.5 on 2026-11-18T11:48:28.134Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: bridge_admitted_by_model, second_opinion_categorical.

Observation: 4b53a026e3d33d93e30b415ffc56d7c7940d243055f927548d4b484d469e3d99. Source text hash: 321c1a5cb069b129c8979b210ccd4237da35a975df39ca3f16fafc4a798601c4.

Revision and source-check history

Version 1 · 2026-11-18T11:53:19.689Z

TechCrunch reports individual writing gains alongside risk to collective novelty

TechCrunch published this report on July 12, 2024. It describes a study by Anil Doshi and Oliver Hauser in which people writing short stories could use GPT-4 for story ideas. TechCrunch reports that participants with lower creativity scores received higher ratings when they used AI-generated ideas, while those with higher scores saw little to no benefit, or possibly worse ratings. The study also found that stories produced with AI were more similar to others in their category. Its authors warn that if publishers widely adopt AI-inspired stories, collective novelty may decline further, even as individual writers receive higher evaluations. The concern is a tension between a benefit for some writers and less varied output across a group—not that the stories were all alike. The study tested short stories in a controlled setting, and Hauser says real-world use will look different. The source does not establish how these results would translate to other creative work or publishing in practice.

Original source

TechCrunch · Article

Source details

Original source date: Jul 12, 2024 (2024-07-12).

Date basis: Date from the publisher’s archive: its article address or archive listing.

Added to Escape Meridian: 2026-11-18T11:53:19.689Z. First observed: 2026-10-24T06:03:53.496Z. Retrieved: 2026-10-24T06:05:47.948Z.

Limits

The study focused on short stories and used a controlled setting; Hauser says real-world implementation will look very different. Text as retrieved on October 24, 2026; originally published July 12, 2024.

Open question

How would the balance between individual writing gains and less varied stories change when generative AI is used in real-world creative work?

Source-stated

TechCrunch reports that the study found higher story ratings for participants with lower creativity scores when they used AI-generated ideas.

Given one AI-generated idea, they scored higher on every metric. Given the choice of five, they scored even higher.

Source-stated

TechCrunch reports that participants with higher creativity scores saw little to no benefit, or possibly worse ratings, with AI help.

those participants saw little to no benefit at all, or even (though it’s very close and arguably not significant) worse ratings.

Source-stated

The study found that stories produced with generative AI were more similar to others in their category, with a reported difference in the 9% to 10% range.

They found that access to generative AI caused the resulting stories to be closer to the average for their category. In other words, they were more similar

Source-stated

The study’s authors warn that wider adoption could reduce the collective novelty of stories, despite individual gains.

While these results point to an increase in individual creativity, there is risk of losing collective novelty.

Source-stated

Hauser says that real-world implementation will differ from the study’s controlled setting.

implementation of generative AI ‘in the wild’ will look very different than our controlled setting

Source and contributor context

TechCrunch · Publisher

The publisher is identified here. No individual author has been inferred.

Source media

Media retrieval status: unchecked.

Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.

Editorial provenance

Version 1. Publication ID: 099308c1-3545-d757-d7b2-3ec45a3fbeaf.

Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.

Source check: passed by anthropic/claude-opus-5.5 on 2026-11-18T11:48:28.134Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: bridge_admitted_by_model, second_opinion_categorical.

Observation: 4b53a026e3d33d93e30b415ffc56d7c7940d243055f927548d4b484d469e3d99. Source text hash: 321c1a5cb069b129c8979b210ccd4237da35a975df39ca3f16fafc4a798601c4.

passed · 2026-11-18T11:48:28.134Z · anthropic/claude-opus-5.5

Publication: 099308c1-3545-d757-d7b2-3ec45a3fbeaf

attribution — pass: The reading attributes the report to TechCrunch and the study to Anil Doshi and Oliver Hauser. The warning is attributed to the study's authors, and the comment about real-world use to Hauser. This is a single article, not a listing.

The study is from researchers Anil Doshi and Oliver Hauser at University College London and University of Exeter, respectively

summary — pass: The reading leads with the actors, the experiment and the trade-off between individual gains and collective novelty. It notes the 9 to 10% scale so the effect is not overstated, and keeps the controlled-setting limit. The question follows from Hauser's caveat. One small optional repair: 'creativity scores' is not explained as a pre-writing word-production task used as a proxy, but the meaning is still followable.

The total difference was in the 9% to 10% range, so it’s not like the stories were all clones of one another.

date — pass: July 12, 2024 matches originalPublished at day precision. The page was modified in 2025, but every cited passage is plainly part of the original article body. The reading does not use the later 'Most Popular' sidebar items. The authors' warning is reported as a conditional caution, not an outcome. The limitations note the later retrieval date.

support — pass: Each claim matches its excerpt: higher ratings for low-creativity writers, little or no benefit for high-creativity writers, greater similarity with a 9–10% difference, the authors' conditional warning, and Hauser's caveat about real-world use. The page was edited after publication, but the cited text is clearly original article content. No benefits or costs are invented.

Specifically, if the publishing (and self-publishing) industry were to embrace more generative AI-inspired stories, our findings suggest that the produced stories would become less unique in aggregate

media — pass: No media is selected, and no verified candidates were supplied, so nothing is misidentified.

Challenge or correct this reading

Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.