Efficiency in Evolution: Leveraging Generative AI for the Revision of Open Textbooks
Tiffani Tijerina and Tamara Powell
In the landscape of open educational resources (OER), the promise of the “living textbook” often clashes with the harsh reality of academic labor. While the open nature of these resources allows for continuous improvement, the manual effort required to maintain currency, accessibility, and stylistic cohesion across hundreds of pages can be overwhelming for faculty authors. This sustainability crisis often results in once-vibrant resources becoming dated and difficult to use.
This case study examines the evolution of Open Technical Communication, 2e, a project in which its authors navigated the OER revision dilemma by integrating generative artificial intelligence (Gen AI) into the production workflow for the second edition of this open textbook. Since its first edition was published in 2015 as Sexy Technical Communication, the project has earned international recognition for the team’s commitment to continuous improvement, including a 2022 Open Education Award for Excellence in Reuse/Remix/Adaptation. However, as the team prepared for a comprehensive revision prior to releasing the book’s second edition in 2025, it became clear that traditional revision methods were insufficient for the scale of the update required.
By adopting a human-in-the-loop philosophy grounded in Jeanne Law’s (2023) rhetorical prompting framework, the project successfully accelerated the revision cycle without sacrificing pedagogical integrity. This chapter explores how generative AI was used to streamline style, modernize examples, and enhance accessibility. It argues that generative AI, when governed by disciplined, rhetorical expertise, offers a sustainable path forward for OER maintenance, enabling authors to transform static materials into dynamic, high-quality resources with unprecedented efficiency.
Background: The Evolution of Open Technical Communication
The project now known as Open Technical Communication originated from a commitment to pedagogical equity and fiscal accessibility. In 2015, Dr. Tamara Powell of Kennesaw State University (KSU) assembled a collaborative team to address a persistent barrier in technical communication pedagogy: the high cost of traditional textbooks. Supported by an Affordable Learning Georgia (ALG) grant, the team sought to replace expensive proprietary materials with a high-quality, open-access alternative. This effort culminated in the 2016 publication of Sexy Technical Communication, a thorough remix of David McMurrey’s foundational Online Technical Writing. While the original title began as a lighthearted internal placeholder, it remained the project’s public identity for its first four years, symbolizing the faculty-led nature of its development.
As the resource matured, its identity shifted toward professional standardization. The team eventually rebranded the text as Open Technical Communication to better align with the institutional and academic environments it served. A pivotal technological shift occurred in 2020 when the text migrated to Manifold, an open-source scholarly publishing platform. Hosted via ALG’s instance of Manifold, called OpenALG, this move transformed the static webpage into a dynamic, interactive ecosystem, introducing robust accessibility features, annotation tools, and an intuitive user interface.
Despite these successes, the project eventually encountered an OER revision plateau. The necessity for a comprehensive structural and content overhaul became apparent, yet the labor required for such a transition–often unpaid and undervalued in traditional tenure tracks–collided with the professional demands of its authors, including doctoral pursuits and the academic job market. The catalyst for the second edition arrived in the summer of 2025 through a small grant from KSU’s Radow College of Humanities and Social Sciences. This grant presented a formidable challenge: a small team of faculty and students was tasked with executing a massive overhaul–including reorganization, content updates, and the addition of new chapters–within three months. This temporal and financial constraint served as the primary driver for exploring Gen AI as a partner in the revision process.
Methodology: The Rhetorical Prompting Framework
The revision of Open Technical Communication was governed by a dual-framework approach designed to maintain academic integrity while maximizing efficiency. Rather than treating generative AI as an autonomous content creator, the team employed it as a sophisticated rhetorical partner. This partnering effort was achieved by synthesizing Dr. Jeanne Law’s Rhetorical Shot Engineering (2023) and Ethical Wheel of Prompting (2025).
Rhetorical Prompt (Shot) Engineering
To ensure the AI’s output aligned with the nuanced demands of technical communication, the team moved beyond simple natural language queries toward disciplined rhetorical prompts. This method involved constructing prompts that provided the AI with details including purpose, audience, tone, genre, style, context, and specific information to include. Prompts were engineered to define essential rhetorical details with specific examples for context such as “re-write this paragraph of text so that the style more closely aligns with the style of the text in the attached document,” “update this example so that it resonates better with contemporary readers,” or “write 3-4 learning outcomes for the attached chapter using only measurable verbs.”
Figure 1
Rhetorical Prompting Method (CC BY-NC-ND 4.0, Jeanne Beatrix Law, 2023)
Ethical Wheel of Prompting
The speed of generative AI necessitates a robust evaluative framework to prevent the propagation of algorithmic bias or factual hallucination. To mitigate these risks, the team utilized the Ethical Wheel of Prompting (Law, 2025) as a continuous feedback loop. The framework required authors to pass every AI-generated output through a “human-in-the-loop” verification process.
Figure 2
Ethical Wheel of Prompting (CC BY-NC 4.0, Jeanne Beatrix Law, 2025)
The wheel functioned as a checklist for three critical areas:
- Transparency and Accuracy: Every AI-assisted revision was cross-referenced against primary technical communication scholarship (as appropriate) to ensure theoretical accuracy. In addition, each chapter went through at least four human read-throughs, including two with the primary authors, Tijerina and Powell, and one each with two expert editors, Michelle Jimmerson and Jenny Rissen.
- Bias Mitigation: Outputs were audited for gendered language, Eurocentric examples, and exclusionary technical jargon, ensuring the text remained globally inclusive.
- Human Agency: The final decision-making power remained with the subject matter experts. If an AI-generated passage failed to capture the rhetorical complexity of a specific topic, such as ethics in technical communication, it was discarded or manually overhauled.
By anchoring the project in these two frameworks, the team transformed the AI from a black box tool into a transparent, rhetorically grounded extension of the faculty’s own expertise.
Core Case Study: Specific AI Applications
The transition from the theoretical framework of Rhetorical Shot Engineering and the Ethical Wheel of Prompting to the practical overhaul of Open Technical Communication was marked by a series of rapid, iterative sprints. Several years before the summer revision process began, Tijerina had carefully examined and drafted a new organizational structure for the text; however, because she was pursuing her PhD at the time, she was unable to act on those plans sooner. So, when the opportunity arose to start the process, the “bones” of the textbook were already firmly in place. The team then prepared a plan for completing their revision rapidly with the assistance of generative AI. The following subsections detail the three primary areas where AI integration proved most transformative.
Content Generation and Structural Coherence
A primary goal for the second edition was to resolve the disjointed nature often found in remixed OER. Many chapters in the first edition lacked strong introductory hooks or smooth transitions between complex technical concepts. To address this, the team used AI to draft introductory paragraphs and synthesizing summaries.
For example, when preparing the chapter on business correspondence, the team combined two chapters from the first edition into one synthesized chapter. In the process, they needed to draft introductory paragraphs and remedy tone and style alignment disparities. Some example prompts used in this process included
“Write a 2-paragraph introduction to the attached sections of text in a chapter on business correspondence for a technical communication textbook.”
“Look at these two sections and suggest shared themes to use for transitions and to improve coherence.”
Using generative AI in this way allowed authors to move from a “blank page” state to a “curatorial” state in minutes, focusing their efforts on refining the AI-assisted draft rather than laboring over structural connective tissue.
Pedagogical and Accessibility Upgrades
Modern OER standards require high levels of accessibility and measurable pedagogical outcomes: areas that are often the most tedious for faculty to update manually.
Measurable Learning Outcomes
Many legacy outcomes in the first edition were basic with little to no detail. In addition, some chapters were combined with others, resulting in unnecessarily long lists of outcomes. To improve learning outcomes in each chapter, the team prompted generative AI to “write 4-5 learning outcomes for the attached chapter using only measurable verbs.” As a result, each chapter in the second edition has a similar number of outcomes that comprehensively preview the content. This consistency was helpful particularly with chapters that combined multiple first edition chapters.
For example, Chapter 2.1, “Business Correspondence and Resumes” and Chapter 2.13, “Memos and Emails” in the first edition carried the following learning outcomes:
Chapter 2.1 Business Correspondence and Resumes
- Summarize the basics of format and style in business correspondence.
- Explain and distinguish between three common types of business letters.
- Explain and apply basic guidelines for resume-writing.
Chapter 2.13 memos and Emails
- identify basic memo and email formats.
- Describe the key differences between basic memo and basic email formats.
- Explain cc, bcc, and attachment functions and when to use them.
After prompting Gen AI for combined and improved outcomes for Chapter 12, “Business Correspondence” in the second edition, the outcomes for the new chapter became the following:
- Differentiate among key forms of business correspondence–including letters, memos, and emails–by identifying their structure, tone, and appropriate use cases.
- Compose a formal business lette using standard components and paply strategies for clarity, organization, and reader-centered communication.
- Write and format professional memos and emails that adhere to workplace conventions and effectively deliver both infomrational and directive content to internal audiences.
- Demonstrate appropriate tone, conciseness, and organization in business messages by revising poorly written examples and paply the “you-attitude,” clear topic sentences, and emphasis strategies.
- Apply techniquees for specific types of correspondence by creating or evaluating examples for professionalism, effectiveness, and reader impact.
Automated Alternative Text
In a textbook heavy on technical illustrations and diagrams, writing descriptive alternative text is a massive undertaking. Using generative AI’s multimodal capabilities, the team uploaded images and requested concise, context-aware alternative text. For example, one prompt used for generating alternative text was “for the attached image in a chapter on business correspondence, draft alternative text of no longer than two sentences for the image.” While every output was human-verified for accuracy, the AI’s ability to describe the spatial layout and components of different types of documents provided a significant head start in meeting Web Content Accessibility Guuidelines (WCAG) 2.1 accessibility standards.
Modernization and Stylistic Alignment
One of the major updates in the second edition of Open Technical Communication was the modernization of technical examples. Technical communication is a field that moves at the speed of industry; a textbook that references transistor radios or mailing a physical resume loses immediate credibility with Gen Z and Gen Alpha students.
The team used Gen AI to update existing examples to more modern topics. For example, one prompt used was “using this example of organizing a paragraph using cause and effect as a guide, change the subject from transistor radios to something more interesting to contemporary students of technical communication.” Prompts like these offered the opportunity to replace decades-old references with examples that are both more relevant and more interesting to current students. For example, in Chapter 10, “Organization and Structure” in the second edition, the team replaced an example of Problem-Methods-Solution paragraph organization that referenced outdated technology (such as CD-ROM drives) with an example that references Star Trek, a long-standing icon of popular culture.
In addition, the team used Gen AI to streamline style and voice across chapters written by different original authors. By feeding the AI a prompt with example text for reference, the team was able to harmonize the second edition’s voice, ensuring that student experience remained consistent from the first page to the last. An example prompt used for style alignment purposes was “re-write this section of text so that the style more closely aligns with the style of the text in the attached document.”
The “PhD-Land” Catalyst
It is critical to note that these applications succeeded only because they were guided by a pre-established roadmap. The “bones” of the project–the new chapter organization and the mapping of content gaps–were developed manually during a period of intense scholarly reflection. During this phase of the project, which the team calls its “PhD-Land” phase because Tijerina was deep in the throes of her dissertation work, the existing chapters were mapped against the structure of actual technical communication courses that previously used the first edition of the textbook. This crucial, manual mapping phase of the project ensured that the AI was never driving the project; it was merely the high-performance engine that allowed the team to cover years’ worth of revision territory in a single summer.
Critical Reflections: What Worked and What Didn’t
The Summer 2025 revision sprint for Open Technical Communication served as a laboratory for testing the boundaries of AI-human collaboration. While the project achieved its goal of a rapid, high-quality overhaul, the process revealed a sharp distinction between where AI thrives and where it falters.
The Successes: Moving from Author to Editor
The most significant win of this approach to revision was the radical reduction of cognitive friction. By using AI to handle the connective tissue of the book-drafting transitions, improving learning outcomes, and updating dated examples–the team was freed from the “blank page” paralysis that often stalls OER updates. The Rhetorical Shot Engineering method essentially allowed the team to act as high-level creative directors. We found that AI is exceptionally good at stylistic mimicry; once provided with a clear voice model, it could replicate the professional yet accessible tone of Open Technical Communication with high accuracy, leaving only the final polish to be completed by the human experts.
The Failures: The “Precision Gap” in Visuals
However, integrating AI with human editors was not without its problems. The most notable failure occurred with AI-generated imagery. While tools like DALL-E excel at conceptual art, they proved inadequate for the precision required for a technical communication textbook. Efforts to generate diagrams or technical illustrations often resulted in visual gibberish: labels with nonsensical text or glaring typos. For a field predicated on clarity and accuracy, these outputs were unusable. We quickly realized that while AI could describe an image (generate alternative text), it could not yet reliably architect an image that met pedagogical standards.
The “PhD-Land” Caveat
Perhaps the most vital reflection is that the AI did not write the second edition; it assembled the second edition based on a human-designed blueprint. The “PhD-Land” phase–the period where the team manually mapped out the new chapter organization and identified content gaps–was the indispensable foundation.
Crucial Lesson: AI is a powerful force multiplier, but if you multiply zero, you still get zero. Without a rigorous, human-led structural plan, the AI-generated content would have lacked the necessary narrative arc and scholarly depth.
Ultimately, the Ethical Wheel of Prompting served as our guardrail. There were moments where the AI’s suggestions were overly “corporate” or lacked the nuance required for discussing workplace ethics. In those instances, human agency was not just a final check. Human agency was the essential corrective that kept the textbook grounded in academic excellence.
Future Directions and Conclusion
The second edition of Open Technical Communication is not merely an updated textbook; it is a proof of concept for a new era of OER development. Moving forward, the team aims to leverage this workflow to build a comprehensive ancillary ecosystem. By applying the same rhetorical prompting frameworks, we plan to generate slide decks, quiz banks, a tutoring chatbot, and other interactive activities–materials that historically required years of dedicated labor. We also see potential in using AI to draft outlines for entirely new chapters on emerging topics like “AI Prompting for Technical Writers,” ensuring the text remains at the cutting edge of the field.
Ultimately, this case study demonstrates that the OER sustainability crisis is no longer an insurmountable hurdle. Gen AI, when governed by subject matter expertise and ethical guardrails, shifts the faculty role from manual laborer to strategic curator. By drastically lowering the barrier to revision, we ensure that open textbooks remain what they were always meant to be: living, breathing resources that evolve alongside the industries they represent. The future of OER is not just open; it is accelerated, allowing us to prioritize student success through materials that are as current as they are accessible.
References
Law, J. B. (2023). Ethical Wheel of Prompting [infographic]. Dr. Jeanne Beatrix Law. https://jeannebealaw.substack.com/p/asking-more-of-ai-what-openais-new
Law, J. B. (2023). Rhetorical Prompting Method. Kennesaw State university Faculty Web. https://facultyweb.kennesaw.edu/jlaw29/Rhetorical_Prompting_Method.png
McMurrey, D. (1995). Online Technical Writing. McMurrey Associates. https://mcmassociates.io/textbook/
Tijerina, T., Powell, T., Jimmerson, M., & Rissen, J. (2025). Open Technical Communication, 2e. OpenALG. http://open-tc.com
Licensing
This chapter, “Efficiency in Evolution: Leveraging Generative AI for the Revision of Open Textbooks” by Dr. Tiffani Tijerina and Dr. Tamara Powell is licensed CC BY-NC-SA 4.0 International and originally written for inclusion in DOERS AI + OER Case Studies. This chapter is based on revision work done on the Open Technical Communication, 2e open textbook and adapted from a presentation given at the Open Arkansas 2025 Conference.
