Large language models and the problem of rhetorical debt
Open accessThe DOI matches its Crossref record; 50 of the 50 references it cites were found in a registry or an index; 3 of the 56 sources its claims cite were checked for support.
Abstract
This article offers broadly useful guidance for society’s adaptation to the omnipresence of generative AI, with implications for every profession and academic discipline that involves writing or coding (recognized by some as a form of writing). Offering an interdisciplinary perspective grounded in the digital humanities, software development and writing across the curriculum, and building on performance historian Christopher Grobe’s research on the role of arts and humanities expertise in AI development, I offer redefinitions of training data and prompt engineering . These essential yet misleading terms obscure the critical roles that humanities-based expertise has played in the development of GPTs and must play in guiding society’s adaptation to generative AI. I also briefly review scholarship on what constitutes “writing” and what it means to teach writing. Next, I reflect on long-terms trends, in professional software development, of code sharing and reliance on automation, and the likely impact of imposing similar practices in professional writing. After identifying the fundamental problem of rhetorical debt and outlining its consequences, I further motivate my argument, in relation to the new economic value of expert writing. This new economic value necessitates a revaluation of the humanities—not only by computer science, the tech industry, and schools and universities, but by humanists themselves.
Overview
CoChat Summary
TL;DR
This conceptual article argues that generative AI’s “training data” is human expertise captured in writing and that “prompt engineering” is really prompt writing, requiring rhetorical judgment. It introduces “rhetorical debt” as the writing analogue of technical debt and calls for revaluing humanities-based writing education as AI increases the economic value of expert writing.
Key findings
- The paper redefines “training data” as human expertise captured in writing that must be continually developed and curated by humans, and “prompt engineering” as prompt writing that details the immediate rhetorical situation—purpose, audience, genre, and local context—which an LLM cannot access without human guidance.
- It identifies “rhetorical debt” as accruing when writers in haste fail to critically edit text to suit audience, purpose, genre, and context, lack the expertise to do so, or uncritically use AI-generated text; this parallels software technical debt and is worsened by automation bias and a “fluency fallacy.”
- The article cites evidence of AI’s dependence on expert human writing: AI companies hired college graduates at 20–40 USD/h to write for and train AI; Taylor & Francis allowed Microsoft to use authors’ work as training data for 10 million USD in the first year; and the 2025 NACE survey ranked written communication skills (77.1%) among the top in-demand skills, behind problem-solving (88.3%) and teamwork (81%).
- It warns that fluent but inappropriate AI output can escape expert attention, citing Stack Overflow’s December 2022 ban on AI-generated text and later human-verified Answer Assistant, Boeing 737 Max 9 disasters that killed 346 people and cost at least $20 billion, Tesla recalls of 2 million vehicles in December 2023 and 2.2 million in February 2024, and hallucination rates as high as 79% in some reasoning systems trained via reinforcement learning.
Methods
- This is a conceptual, interdisciplinary argument grounded in the digital humanities, software development, and writing across the curriculum, building on Christopher Grobe’s research on arts and humanities expertise in AI development.
- The paper reviews scholarship on writing and writing instruction, reflects on long-term trends in professional software development, and develops the concept of rhetorical debt through illustrative examples and secondary sources.
- No empirical sample, dataset, or formal analysis is reported; the author draws on cross-disciplinary experience and cited literature rather than primary data collection.
Limitations
- The paper states that it restricts its consideration of training data to text, excluding images, video, and curated numerical datasets, and that creative writing is outside its scope; it also says reviewing centuries of writing scholarship is beyond its scope.
- No explicit limitations section is present. Evident from the text: the argument is not empirically tested, relies on illustrative cases and secondary literature, and makes predictive claims about future workplace and educational effects that are interpretive rather than experimentally validated.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
3 Valid · 4 Not valid · 44 Not checked · 56 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Now chatbots can effectively generate a text faster than humans can think of, type out, or dictate a similar text (Strijkers and Costa 2011), requiring high-level reading skills to assess that AI output, and high-level editing skills to adjust one's prompt and/or edit the AI output.”
§ Large language models and the problem of rhetorical debt > 3.1 What is writing, anyway?
What the sources say
- Partly supportedcites Strijkers and Costa 2011 matched to “Riding the Lexical Speedway: A Critical Review on the Time Course of Lexical Selection in Speech Production” (2011) sourceExperimental — accuracy not yet measured
The Introduction reports that humans utter about three words per second, supporting only a human speech-rate component of the claim. It does not compare that rate with chatbot text generation or give rates for thinking or typing; the lexical-timing discussion likewise makes no such comparison. (LLM, conf=0.86)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“AI degenerates when trained on its own output (Shumailov et al. 2024).”
§ Large language models and the problem of rhetorical debt > 4.3 The increased economic value of expert writing and advanced writing skills
What the sources say
- Supportedcites Shumailov et al. 2024 matched to “AI models collapse when trained on recursively generated data” (2024) source
The definition of model collapse describes a degenerative process in which generated data pollute the next generation’s training set. The ‘Model collapse in language models’ section reports degraded performance under both tested training regimes, supporting the claim as a broad paraphrase. (LLM, conf=0.92)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As with writing, so with coding: human programmers have a richer sense of context than chatbots.”
§ Large language models and the problem of rhetorical debt > 4.1 Code sharing, software dependencies, and technical debt: efficiency vs. haste
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For one thing, aesthetic experts fail to receive credit or compensation for their intellectual property.4”
§ Large language models and the problem of rhetorical debt > 1.1 AI's dependence on the arts and humanities
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“which the machine cannot access without human guidance, because the LLM is not embedded in the immediate physical, social, cultural, political world in the which the prompting human lives.”
§ Large language models and the problem of rhetorical debt > 3 Prompt "Engineering" is prompt writing
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
20 more claims
Create a free account to see all 107 claims and to verify the 53 references still open.
- It may state two separate claims as one. 39
- Probably not a substantial claim. 61
- The quote only partly states this claim. 2
Citations
Cites
On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445922 · Cited by 6,988
BRUNER, J. S. (1960) · doi:10.4159/9780674028999 · Cited by 6,042
The Uncanny Valley [From the Field]
Mori, M., MacDorman, K., & Kageki, N. (2012) · IEEE Robotics & Automation Magazine · doi:10.1109/mra.2012.2192811 · Cited by 3,542
Licklider, J. C. R. (1960) · IRE Transactions on Human Factors in Electronics · doi:10.1109/thfe2.1960.4503259 · Cited by 1,938
Burke, K. (1969) · doi:10.1525/9780520353237 · Cited by 1,723
Bean, J. C., & Weimer, M. G. (1996) · Cited by 1,448
AI models collapse when trained on recursively generated data
Shumailov, I., Shumaylov, Z., Zhao, Y., Papernot, N., Anderson, R., & Gal, Y. (2024) · Nature · doi:10.1038/s41586-024-07566-y · Cited by 793 · cited 1×
“AI degenerates when trained on its own output (Shumailov et al. 2024).”
Emig, J. A. (1977) · College Composition and Communication · doi:10.2307/356095 · Cited by 753
Does automation bias decision-making?
SKITKA, L. J., MOSIER, K. L., & BURDICK, M. (1999) · International Journal of Human-Computer Studies · doi:10.1006/ijhc.1999.0252 · Cited by 694 · cited 1×
“In the heady weeks after the public release of ChatGPT (on 30 November 2022), some programmers placed undue confidence in the chatbot’s coding advice, due to automation bias—the tendency to place too much trust in machine-generated suggestions (Skitka et al. 1999).”
Weber, S. (2004) · doi:10.4159/9780674044999 · Cited by 611
Elbow, P. (1998) · doi:10.1093/oso/9780195120165.001.0001 · Cited by 548
Lee, H.-P. (., Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R., & Wilson, N. (2025) · Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3713778 · Cited by 490 · cited 1×
“Already, research suggests that when knowledge workers place “higher confidence” in AI, they do less critical thinking (Lee et al. 2025).”
Becker, B. A., Denny, P., Finnie-Ansley, J., Luxton-Reilly, A., Prather, J., & Santos, E. A. (2023) · Proceedings of the 54th ACM Technical Symposium on Computer Science Education V. 1 · doi:10.1145/3545945.3569759 · Cited by 401 · cited 1×
“Like humanities faculty struggling to adapt to teaching writing amid the omnipresence of chatbots, computer science educators have been trying to figure that out (Becker et al. 2023; Zastudil et al. 2023).”
Hinds, P. J. (1999) · Journal of Experimental Psychology: Applied · doi:10.1037/1076-898x.5.2.205 · Cited by 397
Dissociating language and thought in large language models
Mahowald, K., Ivanova, A. A., Blank, I. A., Kanwisher, N., Tenenbaum, J. B., & Fedorenko, E. (2024) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2024.01.011 · Cited by 345
Modern code review: a case study at google
Sadowski, C., Söderberg, E., Church, L., Sipko, M., & Bacchelli, A. (2018) · Proceedings of the 40th International Conference on Software Engineering: Software Engineering in Practice · doi:10.1145/3183519.3183525 · Cited by 280 · cited 1×
“Code review has evolved to rely on more and more automation, yet human programmers have continued to play a crucial role, as with Google’s code review process (Sadowski et al. 2018).”
Classical Rhetoric for the Modern Student
Winterowd, W. R., & Corbett, E. P. J. (1966) · College Composition and Communication · doi:10.2307/354677 · Cited by 206
A Large-Scale Survey on the Usability of AI Programming Assistants: Successes and Challenges
Liang, J. T., Yang, C., & Myers, B. A. (2023) · arXiv · doi:10.48550/arxiv.2303.17125 · Cited by 206 · cited 1×
“…code: concerns about code quality (including defects, security issues, and performance problems), difficulties in understanding and debugging generated code, time wasted modifying AI-generated solutions, and privacy concerns about tools having access to their proprietary code (Liang et al. 2024).”
A Large-Scale Survey on the Usability of AI Programming Assistants: Successes and Challenges
Liang, J. T., Yang, C., & Myers, B. A. (2024) · Proceedings of the IEEE/ACM 46th International Conference on Software Engineering · doi:10.1145/3597503.3608128 · Cited by 203 · cited 1×
“…code: concerns about code quality (including defects, security issues, and performance problems), difficulties in understanding and debugging generated code, time wasted modifying AI-generated solutions, and privacy concerns about tools having access to their proprietary code (Liang et al. 2024).”
Understanding Computer Programming as a Literacy
Vee, A. (2013) · Literacy in Composition Studies · doi:10.21623/1.1.2.4 · Cited by 168
A Critical Analysis of the Largest Source for Generative AI Training Data: Common Crawl
Baack, S. (2024) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3630106.3659033 · Cited by 152 · cited 1×
“As we know, the LLMs that underlie chatbots excel at engaging in and simulating humanlike dialog because they were trained—using deep machine learning techniques that imitate the neural networks of the human brain—on a massive amount of text written by humans, notably from Common Crawl (Baack 2024).”
Generative AI in Computing Education: Perspectives of Students and Instructors
Zastudil, C., Rogalska, M., Kapp, C., Vaughn, J., & MacNeil, S. (2023) · 2023 IEEE Frontiers in Education Conference (FIE) · doi:10.1109/fie58773.2023.10343467 · Cited by 136
AI and the problem of knowledge collapse
Peterson, A. J. (2025) · AI & Society · doi:10.1007/s00146-024-02173-x · Cited by 127 · cited 1×
“To predict and mitigate the negative “impact[s] of widespread … adoption” of AI on the “future of human thought, information-seeking, and knowledge” (Peterson 2025), we must focus on that challenge.”
Execution and pauses in writing narratives: Processing time, cognitive effort and typing skill
Alves, R. A., Castro, S. L., & Olive, T. (2008) · International Journal of Psychology · doi:10.1080/00207590701398951 · Cited by 114 · cited 1×
“These recent writing technologies began to approach the speed of human thought in the drafting process (Alves et al. 2008).”
Strijkers, K., & Costa, A. (2011) · Frontiers in Psychology · doi:10.3389/fpsyg.2011.00356 · Cited by 106 · cited 1×
“Now chatbots can effectively generate a text faster than humans can think of, type out, or dictate a similar text (Strijkers and Costa 2011), requiring high-level reading skills to assess that AI output, and high-level editing skills to adjust one’s prompt and/or edit the AI output.”
Welsh, M. (2023) · Communications of the ACM · doi:10.1145/3570220 · Cited by 93 · cited 1×
“(Welsh 2023) A major misconception in Welsh’s vision of this future (not to diminish problems with the automation of commercial flight, distribution of electricity, and governance) is that he assumes supervising and educating are inferior to programming, and perhaps easier (“relegate,” according to…”
A Theory of Discourse: The Aims of Discourse
Freeman, R. E., & Kinneavy, J. L. (1973) · College Composition and Communication · doi:10.2307/356519 · Cited by 89 · cited 1×
“…self-expressive and personal development potential of writing (Elbow 1973, Macrorie 1970, 1985, Murray 1968, 2004, Ward 1994), and the rhetorical dimensions of writing, with specific attention to audience (Burke 1950; Corbett 1965; Perelman and Oblrechts-Tyteca 1958; Murphy 1972; Kinneavy 1971).”
Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence
Hankwitz, M. B. (2022) · Leonardo · doi:10.1162/leon_r_02206 · Cited by 72
Rhetorical Code Studies: Discovering Arguments in and around Code
Brock, K. (2018) · University of Michigan Press eBooks · doi:10.3998/mpub.10019291 · Cited by 69
Code Review Automation: Strengths and Weaknesses of the State of the Art
Tufano, R., Dabić, O., Mastropaolo, A., Ciniselli, M., & Bavota, G. (2024) · IEEE Transactions on Software Engineering · doi:10.1109/tse.2023.3348172 · Cited by 48 · cited 1×
“Recent assessments of the accuracy of automated code review report quite mixed results, noting that ChatGPT lacks skill in understanding the larger context of a program, as opposed to a small bit of a code, as well as with making recommendations and implementing changes (Tufano et al. 2024).”
Literacy, Ideology, and Dialogue: Towards a Dialogic Pedagogy
Ward, I. (1995) · College Composition and Communication · doi:10.2307/358734 · Cited by 39 · cited 2×
“…Emig 1971; Murray 1972), the self-expressive and personal development potential of writing (Elbow 1973, Macrorie 1970, 1985, Murray 1968, 2004, Ward 1994), and the rhetorical dimensions of writing, with specific attention to audience (Burke 1950; Corbett 1965; Perelman and Oblrechts-Tyteca…”
Becker, B. A., Denny, P., Finnie-Ansley, J., Luxton-Reilly, A., Prather, J., & Santos, E. A. (2022) · arXiv · doi:10.48550/arxiv.2212.01020 · Cited by 36
Large language models (LLMs) and the institutionalization of misinformation
Garry, M., Chan, W. M., Foster, J., & Henkel, L. A. (2024) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2024.08.007 · Cited by 33 · cited 1×
“Yet these tools are designed to sound confident, and we are worrisomely prone to trust confident fluency, in AI and people (Garry et al. 2024); the “con” in “con man” stands, of course, for confidence (see Herman Melville’s 1857 novel The Confidence Man ).”
Peer and AI Review + Reflection (PAIRR): A human-centered approach to formative assessment
Sperber, L., MacArthur, M., Minnillo, S., Stillman, N., & Whithaus, C. (2025) · Computers and Composition · doi:10.1016/j.compcom.2025.102921 · Cited by 29
Self-citationVee, A., Laquintano, T., & Schnitzler, C. (2023) · doi:10.37514/twr-j.2023.1.1.02 · Cited by 29
AI hallucinations are a feature of LLM design, not a bug
Dumit, J., & Roepstorff, A. (2025) · Nature · doi:10.1038/d41586-025-00662-7 · Cited by 28 · cited 2×
“Such AI-generated text, in contrast to obvious slop, is not guilty of factual inaccuracies or so-called hallucinations (Dumit and Roepstorff 2025), but includes content, and makes stylistic choices, which are inappropriate or inadequate to our immediate context, purpose and audience.”
Programming is Writing is Programming
Hermans, F., & Aldewereld, M. (2017) · International Conference on Art, Science, and Engineering of Programming (<Programming>) · doi:10.1145/3079368.3079413 · Cited by 27 · cited 1×
“…useful guidance for society’s adaptation to the omnipresence of generative AI, with implications for every profession and academic discipline that involves writing or coding, recognized by some as a form of writing (Brock 2019; Hermans and Aldewereld 2017; Juszkiewicz et al. 2019; Vee 2013).”
Rhetorical Machines: Writing, Code, and Computational Ethics
JONES, J. H., & Hirsu, L. (2019) · ENLIGHTEN (Jurnal Bimbingan dan Konseling Islam) · Cited by 25
AI and the University as a Service
Kirschenbaum, M., & Raley, R. (2024) · PMLA/Publications of the Modern Language Association of America · doi:10.1632/s003081292400052x · Cited by 20 · cited 1×
“For an inherently fluid technological situation, with new applications and implementations seemingly by the day, and the models themselves continuously evolving, there can be no stable and singular path forward (Kirschenbaum and Raley 2024).”
Glynn, A. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-32729-1_19 · Cited by 20
Zheng, K. (2025) · Linguistics and Education · doi:10.1016/j.linged.2025.101406 · Cited by 16 · cited 1×
“In English-dominant academic and corporate cultures where grammar is policed and accented English is implicitly or explicitly disparaged (Zheng 2025), we have come to associate expertise with normative grammar, idiomatic syntax and vocabulary and slick delivery.”
Forecasting technical debt evolution in software systems: an empirical study
Aversano, L., Bernardi, M. L., Cimitile, M., Iammarino, M., & Montano, D. (2022) · Frontiers of Computer Science · doi:10.1007/s11704-022-1541-7 · Cited by 15 · cited 1×
“TD [Technical Debt] describes the financial implications of choosing to develop the software differently than how [it] should be done” (Aversano et al. 2023).”
AI and its consequences for the written word
Hellström, T. (2024) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2023.1326166 · Cited by 15 · cited 1×
“With remarkable short-sightedness, some also argue that “the valuation of the ability to write” will inevitably diminish in the age of AI (Hellström 2024).”
Robbins, S. (2025) · AI & Society · doi:10.1007/s00146-024-02169-7 · Cited by 13
Generative AI in Computing Education: Perspectives of Students and Instructors
Zastudil, C., Rogalska, M., Kapp, C., Vaughn, J., & MacNeil, S. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2308.04309 · Cited by 11
A Short History of Writing Instruction
(2020) · doi:10.4324/9781003020899 · Cited by 6 · cited 2×
“…manual writing technolo-gies—actually requiring the use of our hands—has accelerated since the late nineteenth century, first with the revolutionary invention of typewriters, which democratized access to a technology available since the invention of the printing press (Murphy and Thaiss 2020, 5–7).”
OpenAI's Pharmacy? On the Phaedrus Analogy for Large Language Models
Kitzinger, C. (2024) · Critical AI · doi:10.1215/2834703x-11205203 · Cited by 5 · cited 1×
“Thus consumption of too much sloppy AI-generated text [which I purposefully do not refer to as “writing” (see Kitzinger 2024; Vee 2023)] may rot the brain.”
Writing Instruction in U.S. Colleges and Schools
Gold, D., & Hammond, J. (2020) · doi:10.4324/9781003020899-8 · Cited by 4 · cited 2×
“Research on and theories of writing instruction since the 1950s have further illuminated the activity of writing, as Gold and Hammond (2020) detail.”
Vee, A. (2024) · Critical AI · doi:10.1215/2834703x-11205182 · Cited by 3 · cited 1×
“Vee has argued that, by releasing GPTs without adequate pre-mortem analysis, tech companies have effectively imposed a huge burden of technical debt on the public (Vee 2024).”
Goodlad, L. M. E. (2023) · Critical Inquiry · doi:10.1086/722404 · Cited by 2
Research works printed but not linked: 12
- 7. Battersby M (2024) Academic authors "shocked" after Taylor & Francis sells access to their research to Microsoft AI. The Book- seller. https:// www. thebo oksel ler. com/ news/ acade mic-autho rs-shock ed-after-taylo r--franc is-sells-access-to-their-resea rch- to-micro soft-ai. Accessed 4 Aug 2024
- 11. Bitzer LF (1968) The rhetorical situation. Philos Rhet 1(1):1-14. http:// www. jstor. org/ stable/ 40236 733
- 21. Dyos S (2025) Sam Altman says the kid he's expecting soon will never be smarter than AI. Fortune. https:// fortu ne. com/ 2025/ 01/ 18/ sam-altman-openai-kid-smart er-than-agent ic-ai-abili ty- skills/. Accessed 25 Jan 2025 Ecclesiastes 1:9. The New Oxford Annotated Bible (3rd ed.) (450~150 BCE, 2001 CE).
- Ecclesiastes 1:9. The New Oxford Annotated Bible (3rd ed.) (450~150 BCE, 2001 CE). Coogan M et al (eds) Oxford University Press
- 29. Glynn A (2024) Suspected undeclared use of artificial intelligence in the academic literature: an analysis of the Academ-AI Dataset. arXiv.org. https:// arxiv. org/ abs/ 2411. 15218
- 33. Gray K, Koncz A (2025) The attributes employers look for on new grad resumes-and how to showcase them. National Association of Colleges and Employers. https:// www. nacew eb. org/ about-us/ press/ the-attri butes-emplo yers-look-for-on-new-grad-resum es- and-how-to-showc ase-them. Accessed 20 Mar 2025
- 63. Murphy JJ (ed) (1972) A synoptic history of classical rhetoric. Random House, New York
- 67. Murray DM (1972, 2003) Teach writing as a process not product. In: Villanueva V (ed) Cross-talk in comp theory: a reader, 2nd edn. National Council of Teachers of English, pp 3-6 Oxford Word of the Year: Our 2024 Shortlist (2024) Oxford Univer- sity Press. https:// corp. oup. com/ word-of-the-year/# short list-2024. Accessed 28 Jan 2025
- Oxford Word of the Year: Our 2024 Shortlist (2024) Oxford University Press. https://corp.oup.com/word-of-the-year/#shortlist-2024. Accessed 28 Jan 2025
- 69. Wilkinson J and Weaver P. Univer- sity of Notre Dame Press
- 74. Rohman DG (1965) Pre-writing: the stage of discovery in the writing process. Coll Compos Commun 16(2):106-112
- 81. Tenen DY (2024) Literary theory for robots: how computers learned to write. W.W. Norton, New York
Cited by
MacArthur, M. J., Minnillo, S., Sperber, L., Whithaus, C., & Stillman, N. (2025) · Frontiers in Communication · doi:10.3389/fcomm.2025.1615752 · Cited by 8
Self-citation“Generative AI” and Writing in Higher Education: Editor's Introduction
MacArthur, M. J. (2025) · Critical AI · doi:10.1215/2834703x-12095937 · Cited by 2
Self-citationLanguage and time in education: thinking in the age of generative artificial intelligence
Cui, S. (. (2026) · AI & Society · doi:10.1007/s00146-026-03148-w
Muthmainnah, M., Cardoso, L., Yakin, A. M. A., Asha, L., Baihaqi, Y., Apriani, E., & Supardan, D. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-7097-2.ch009
PostItFlow: An Early Study on Agent-Based Workflow for Enhancing and Visualizing User Stories
Weerakoon, O., Rytilahti, J., Mäkilä, T., Kaila, E., & Oyedeji, S. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-12089-2_42
Metadata
- Authors
- MacArthur, M.
- Affiliations
- University of California, Davis (US)
- Corresponding author
- MacArthur, Marit
- Venue
- AI & SOCIETY
- Published
- 2025-06-25
- Volume
- 40
- Issue
- 8
- Pages
- 6425-6438
- ISSN
- 0951-5666 (ISSN-L)
- 1435-5655
- OpenAlex
- W4411661084
- Cited by
- 8
- References
- 44
- Citations by year
- 2026: 5 · 2025: 3
- FWCI
- 8.28
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor)
- Locations
- AI & Society (publishedVersion, cc-by)
- SSRN Electronic Journal (acceptedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Social Sciences › Arts and Humanities › Literature and Literary Theory › Digital Humanities and Scholarship
- Social Sciences › Arts and Humanities › History and Philosophy of Science › Cybernetics and Technology in Society
- Keywords
- generative AI, large language models, training data, prompt engineering, writing instruction, professional software development, code sharing, automation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-09-30, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-09-30 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.