How authorship is made and named: a meta-aggregator / Wordcraft Writers Workshop
Wordcraft Writers Workshop
https://wordcraft-writers-workshop.appspot.com/
empirical comparatorMachine participation in authorshipAttribution, credit and identity infrastructurehistorical
The archive's reading. The workshop publishes each story under its writer's name, keeps the writers identified by initials in the paper for the sake of their published work, and records workflows ranging from largely generated stories to stories mostly written by their authors; writers feared 'reputational harm' from unknown sources of suggestions. The archive reads it as machine participation declared at the level of the tool and the cohort, with authorship held by the named writer, and as a provider-run record whose own notice separates Google from the content.
| Relation to this census | empirical comparator |
|---|---|
| Record type | project |
| The question it answers | contribution and authorial function: how named professional writers use a LaMDA-based editor and keep authorship of the stories, as the workshop records it |
| Kind | in its own words, 'a collaboration between Google's PAIR and Magenta teams, and 13 professional writers'; a published story collection, a project site and a whitepaper |
| Operator | Google Research: the PAIR and Magenta teams; conducted by Andy Coenen and Ann Yuan (PAIR), Daphne Ippolito (Brain Research, working with Magenta) and Sehmon Burnam (Magenta) |
| Evidence | primary text; research paper; empirical study |
| Unit of record | a story of 1,000-1,500 words by a named writer, made with Wordcraft over eight weeks, with the writer's journal and two interviews |
| What it records | thirteen stories by thirteen commissioned professional writers using Wordcraft, a LaMDA-powered text editor, and the researchers' qualitative findings from their journals and interviews |
| Whose authorship it keeps | the thirteen writers' authorship of their stories |
| Under what name | each story stands under its writer's name (Robin Sloan, Diana Hamilton, Eugenia Triantafyllou, Ken Liu, Allison Parrish and others); the paper refers to participants by initials and gives its reason for naming them: 'because each writer's perspective is best understood within the context of their background and prior published work'; LaMDA is credited only as the model under the tool; one writer named the AI 'Rain Man' in a diary |
| Who holds it | Google, on its research site, with the notice 'Google does not endorse the content of any of the stories'; participants consented to publication of their stories and use of their feedback, and were compensated; site terms on a page not saved |
| Authorial function and responsibility | the site states the writers 'have used Wordcraft along with their own creative vision and have not authored these stories under any sort of explicit guidance'; the paper records the share of generated text by writer: some made stories 'largely formed around generated text (AP, DH, JM, MT, AW)', for others 'the bulk of the story text was written by the authors themselves (KL, WT, NG)', and one siloed generated text into set sections (RS); the per-sentence division is not published |
| Method | a commissioned eight-week workshop: writers given no instruction or tutorial, asked for a story and a journal on set questions, interviewed twice for 45 minutes; two paper authors coded journals and interview notes with 10 high-level codes |
| Coverage | geography: 'professional English-language writers from around the world'; period: eight weeks; the workshop's dates are not stated in the saved sources |
| Count | value: 13 writers, 13 stories, 8 weeks; as_of: 2026-11-01 (site read) |
| Holdings | thirteen stories, each under its writer's name; a 'learn more' page of findings; the whitepaper, Creative Writing with an AI-Powered Writing Assistant: Perspectives from Professional Writers |
| Stated limits | 'Wordcraft was designed by engineers at Google with limited feedback from professional writers during its development.'; it cannot separate interface weaknesses from limits of the underlying model; it does not test whether the findings extend to other model families; 'our study was very small with only 13 participants, a sample that was biased toward people with pre-existing opinions on the role of AI in creative writing' |
| Data access | free public website with the stories, the findings page and the whitepaper PDF |
| Licence | not stated in the saved sources (a privacy & terms page is linked and unsaved) |
| Lifecycle | state: historical; detail: a completed eight-week workshop; the site answered on 2026-11-01; no date for the workshop in the saved sources (the paper cites work of 2022) |
| Version and time | workshop: eight weeks, date not stated; read: 2026-11-01 |
What its record can settle
- how thirteen named professional writers report using a LaMDA-based editor, and how much generated text each workflow admitted, by initials
- under whose names the stories are published and on what consent
- what the writers reported on voice, style and the unknown origin of suggestions
What it cannot settle
- which sentences of a given story LaMDA generated
- how writers outside this commissioned, provider-organized cohort work
- the source of any suggestion in LaMDA's training data
Kept apart
- NOT authored by LaMDA: the stories stand under their writers' names
- NOT a neutral census of writers: a provider-organized, commissioned cohort of 13
- NOT a record of which text was generated: workflows are reported by writer, without per-sentence marking
- NOT endorsed by Google in content: the site's own notice
Relations
- neighbour: rerites (Allison Parrish, a workshop writer, writes a response essay in ReRites)
Evidence
- self_description (the resource's own statement):
The Wordcraft Writers Workshop is a collaboration between Google's PAIR and Magenta teams, and 13 professional writers from a diverse set of creative writing backgrounds.
https://wordcraft-writers-workshop.appspot.com/ · read 2026-11-01 · http 200 · sha256 963e8e672898885a… - aim (the resource's own statement):
Together we explore the limits of co-writing with LaMDA and foster an honest and earnest conversation about the rapidly changing relationship between technology and creativity.
https://wordcraft-writers-workshop.appspot.com/ · read 2026-11-01 · http 200 · sha256 963e8e672898885a… - story_first (the resource's own statement):
Author's Note Robin Sloan
https://wordcraft-writers-workshop.appspot.com/ · read 2026-11-01 · http 200 · sha256 963e8e672898885a… - story_parrish (the resource's own statement):
Further Frankenstein Variations Allison Parrish
https://wordcraft-writers-workshop.appspot.com/ · read 2026-11-01 · http 200 · sha256 963e8e672898885a… - story_last (the resource's own statement):
Some Crimes Aaron Winslow
https://wordcraft-writers-workshop.appspot.com/ · read 2026-11-01 · http 200 · sha256 963e8e672898885a… - writers_notice (the resource's own statement):
The writers featured here have used Wordcraft along with their own creative vision and have not authored these stories under any sort of explicit guidance or instruction.
https://wordcraft-writers-workshop.appspot.com/ · read 2026-11-01 · http 200 · sha256 963e8e672898885a… - endorse (the resource's own statement):
Google does not endorse the content of any of the stories contained on this website.
https://wordcraft-writers-workshop.appspot.com/ · read 2026-11-01 · http 200 · sha256 963e8e672898885a… - tool (the resource's own statement):
Wordcraft is a tool built by researchers at Google PAIR for writing stories with AI. The application is powered by LaMDA
https://wordcraft-writers-workshop.appspot.com/learn · read 2026-11-01 · http 200 · sha256 60c4489ee0b1379e… - cohort (the resource's own statement):
The workshop cohort consisted of 13 professional writers, who were given 8 weeks to use Wordcraft.
https://wordcraft-writers-workshop.appspot.com/learn · read 2026-11-01 · http 200 · sha256 60c4489ee0b1379e… - writers_world (the resource's own statement):
we brought together 13 professional English-language writers from around the world to use Wordcraft to write stories
https://wordcraft-writers-workshop.appspot.com/learn · read 2026-11-01 · http 200 · sha256 60c4489ee0b1379e… - no_instruction (the resource's own statement):
We explicitly didn't give writers any instructions regarding how they should use the tool or how much of their final story text should be directly generated by Wordcraft.
https://wordcraft-writers-workshop.appspot.com/learn · read 2026-11-01 · http 200 · sha256 60c4489ee0b1379e… - conducted (the resource's own statement):
The Wordcraft Writers Workshop was conducted by Andy Coenen and Ann Yuan, research engineers on the PAIR team, Daphne Ippolito, a research scientist working with Magenta , and Sehmon Burnam, a product manager at Magenta .
https://wordcraft-writers-workshop.appspot.com/learn · read 2026-11-01 · http 200 · sha256 60c4489ee0b1379e… - paper_title (the resource's own statement):
Creative Writing with an AI-Powered Writing Assistant: Perspectives from Professional Writers
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - commissioned (the resource's own statement):
we commissioned 13 professional, published writers from a diverse set of creative writing backgrounds to craft stories using Wordcraft
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - voice_finding (the resource's own statement):
NLG technologies struggle to preserve style and authorial voice, and they lack deep understanding of story contents.
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - initials (the resource's own statement):
We choose to refer to each participant by their initials (see Table 2) rather than anonymizing them because each writer's perspective is best understood within the context of their background and prior published work.
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - story_length (the resource's own statement):
Participants were asked to write a story of 1,000-1,500 words using Wordcraft.
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - interviews (the resource's own statement):
We also conducted two 45 minute interviews with each participant
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - consent (the resource's own statement):
Participants consented to their stories being published in a digital literary magazine of human-AI collaborative stories, and to the use of their feedback in the journals and interviews in this whitepaper. Participants were compensated for participating in the study.
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - coding (the resource's own statement):
Two authors of this paper coded the participants’ journals as well as the authors’ notes from the two interviews.
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - codes (the resource's own statement):
Our high-level code set included 10 codes
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - generated_share (the resource's own statement):
Several participants took the workshop as a challenge to produce a story that was largely formed around generated text (AP, DH, JM, MT, AW).
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - authored_share (the resource's own statement):
the bulk of the story text was written by the authors themselves (KL, WT, NG).
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - siloed (the resource's own statement):
some participants siloed out specific sections of their stories to which generated text could be included without the author needing to cede too much creative control to Wordcraft (RS).
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - origin_concern (the resource's own statement):
professional writers face reputational harm if their work is found to have plagiarized, especially if the copied material came from smaller, more obscure writers.
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - rain_man (the resource's own statement):
I finally settled on Rain Man
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - limit_design (the resource's own statement):
Wordcraft was designed by engineers at Google with limited feedback from professional writers during its development.
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - limit_disambiguate (the resource's own statement):
it is not easily possible to disambiguate weaknesses stemming from Wordcraft’s interface and implementation from inherent limitations of the underlying language model model backing Wordcraft.
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - limit_untested (the resource's own statement):
but this study does not test this hypothesis.
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4… - limit_sample (the resource's own statement):
our study was very small with only 13 participants, a sample that was biased toward people with pre-existing opinions on the role of AI in creative writing.
https://wordcraft-writers-workshop.appspot.com/paper/wordcraft_writers_workshop_paper.pdf · read 2026-11-01 · http 200 · sha256 3618c2818fd597f4…
Unchecked
the thirteen stories themselves (listed on the home page; the story pages are unsaved); the privacy & terms page; the workshop's dates (not stated in the saved sources); the paper's appendix of participant biographies was not read in full
Cite: Lee Sharks, ed. "Wordcraft Writers Workshop." How authorship is made and named: a meta-aggregator, leesharks.com, compiled 2026-11-01. https://www.leesharks.com/apparatus/aggregators/wordcraft