Prompt Trails in the Digital Wilderness: Mapping ChatGPT Prompt Rhetorics on X to Build Critical Digital Literacies
Imagine wandering through a vast, shimmering AI landscape where people trade tips, tricks, and templates for talking to chatbots like ChatGPT. That’s basically what Gupta and Shivers-McNair set out to explore: how conversations about prompts on X (the platform formerly known as Twitter) reveal not just how we prompt machines, but how those prompts shape our thinking, teaching, and sense of literacy in a world infused with AI. Their study is less about “how to make better prompts” and more about understanding the culture, promises, and risks wrapped up in prompt writing as a new kind of social practice. In other words, they map the rhetorics of prompting to help educators nurture critical, ethical, and flexible AI literacies.
In this blog post, I’ll walk you through the key ideas in plain language, connect them to what teachers and students can actually do in classrooms or workshops, and offer practical takeaways you can start using today. No heavy jargon required.
Wayfinding in a Shifting Terrain: What this study tries to do
At its core, the researchers use a “wayfinding” lens. Think of wayfinding as a traveler’s toolkit for navigating unfamiliar terrain: orient yourself, chart rough routes, map what you learn, and make sense of what comes next. In the context of digital literacies, this means recognizing that learning to write prompts for AI isn’t a straight-line journey. It’s iterative, context-dependent, and shaped by human and machine interactions, evolving platforms, and broader social forces.
Gupta and Shivers-McNair argue that studying how people talk about prompting on social media offers a rich way to understand emerging AI literacies. Rather than prescribing a single “correct” way to prompt, they propose examining the discourses around prompting to help students (and teachers) engage thoughtfully with AI as a tool for writing, learning, and thinking.
From Big Data to Close Reading: How the researchers worked
Here’s the high-level journey they describe, translated into plain language:
Data collection: They started with a large pool of posts about ChatGPT and writing on X. In their broader corpus, they gathered about 258,661 posts from late 2022 to mid-2023. Then they zeroed in on a subset—roughly 32,000 posts that were explicitly about prompts or prompt writing.
Mixed methods approach: They blended computational methods with qualitative analysis. First, they used descriptive statistics and a machine-learning technique called topic modeling (BERTopic) to surface patterns in the large dataset. Then they did close, human-centered reading of a smaller, carefully chosen set of posts to interpret what these patterns meant.
Iterative, reflexive process: The study emphasizes that data collection and analysis aren’t neutral processes. The researchers acknowledge their own positionalities and the evolving nature of social media platforms. Their approach is intentionally iterative: let the data guide you, then bring in theory and context, then circle back to bigger questions.
Focus on five emergent themes: Through this process, they identify five themes that illuminate how prompting discourse is shaping AI literacies: (1) areas of communication impacted by prompt writing, (2) micro-literacy resources shared for prompt writing, (3) market rhetoric shaping prompt writing, (4) rhetorical characteristics of prompts, and (5) definitions of prompt writing.
Five themes, unpacked for everyday readers
Theme 1. Areas of communication impacted by prompt writing
- What it looks like: Posts discuss prompts affecting a wide range of writing and communication tasks—not just “academic writing” but also marketing copy, emails, journalism, storytelling, and even coding. In their top-100 posts about prompts, more than half touched on how prompting could change various kinds of communication.
- Why it matters: If prompts can influence emails, ads, or code, then prompting practices aren’t just a tech curiosity; they shape real-world workflows and power dynamics in workplaces, classrooms, and online discourse.
- Takeaway for teaching: When you teach prompting, you’re not just teaching a trick for a chatbot. You’re teaching students to recognize where AI-assisted writing fits into different genres and communities, from marketing to academic work to personal reflection.
Theme 2. Micro-literacy resources shared for prompt writing
- What it looks like: Many top posts share bite-sized, easily digestible resources—prompt databases, tips, “tweetorials,” and short guides. The study calls these micro-literacies: small, easily shareable bits of knowledge that help people prompt more effectively.
- Why it matters: These micro-resources shape how beginners learn to talk to AI and how more experienced users teach others. They also reflect a culture of fast sharing and community curation.
- Takeaway for teaching: Encourage students to curate their own prompt libraries and remix resources critically. Help them think about what makes a prompt database useful (clarity, context, ethical considerations) and how to adapt prompts to different audiences and goals.
Theme 3. Market rhetoric shaping prompt writing
- What it looks like: About half of the top posts engage in market-style rhetoric. They pitch prompts as valuable skills, promise productivity boosts, or frame prompting as a competitive advantage. Some posts lean on “deficit” narratives (you’re currently poor at prompting, here’s the fix) and “engagement” economics (free resources in exchange for follows, likes, etc.).
- Why it matters: This isn’t just about marketing; it signals how the broader economy around AI tools can subtly steer what gets taught, learned, and valued. It also flags the ethics of promotion—what is being sold, and who benefits.
- Takeaway for teaching: Bring a critical lens to the marketing around prompts. Teach students to ask who benefits from a given resource, what assumptions about work and productivity are embedded in these messages, and how to evaluate prompts beyond “how flashy is this” to “how fair and accurate is this for my purpose and audience.”
Theme 4. Rhetorical characteristics of prompts
- What it looks like: Prompts themselves carry rhetorical features. People discuss length (longer prompts sometimes seen as better), tone, templates, data sources, correctness, and even the social niceties of addressing the AI (please/thank you). There’s a sense that prompting sits in an “in-between” space—between a conversation and a structured instruction.
- Why it matters: If prompts are rhetorical acts with tone, audience awareness, and purpose, then teaching prompting becomes a class in rhetorical craft—how to shape outputs for specific audiences and contexts.
- Takeaway for teaching: Use real-world examples to show how tiny changes in tone, audience, or context change the results. Experiment with both conversational prompts and more structured templates, then discuss which approach fits which tasks.
Theme 5. Definitions of prompt writing
- What it looks like: Some posts define prompting as a usability enhancement, others frame it as a skill, a way of talking to machines, or a method of thinking. The definitions reveal a spectrum: prompts as practical tools for better results, and prompts as a gateway to new ways of thinking about language and agency.
- Why it matters: Definitions shape how people approach prompting in practice and in the classroom. If prompting is seen mostly as a “skill,” the emphasis might be on performance; if it’s seen as a “thinking” practice, it invites reflection on human-AI collaboration, ethics, and representation.
- Takeaway for teaching: Encourage students to articulate their own definitions of prompting for different tasks. Use localization and context, as the authors suggest, so prompts feel usable in real communities (academic, professional, or public-facing).
Bringing it to classrooms and real life: pedagogical implications and practical steps
What Gupta and Shivers-McNair offer is not a one-size-fits-all manual, but a flexible framework—what they call a wayfinding-inspired pedagogy—that invites teachers to co-explore prompting with students. Here are some concrete ways to translate their ideas into teaching or workshop practice:
Start with the big picture, then zoom in: Begin with a discussion of where prompting sits in everyday communication (Theme 1) and why it matters beyond “getting a nice essay from a chatbot.” Then move into the more granular resources (Theme 2) and the rhetorical stakes (Themes 3–5).
Use Ranade et al.’s rhetorical prompting formula as a gym for thinking: Prompt ¬ (audience, genre, purpose, subject, context, exigence, writer). Have students analyze an existing prompt (or a sample prompt from a database) using this formula. Then task them with creating their own prompts tailored to a specific audience and purpose. This helps connect theory to practice.
Treat micro-literacies as both tools and lenses: Collect prompts from a class, curate a mini-prompt database, and remix prompts for different communities. Pair this with reflective writing on why certain prompts work in particular contexts and what values are encoded in the choices.
Examine market rhetoric critically: Have students map who benefits from a given prompt resource, what promises are being made, and what ethical considerations might be overlooked in “free” resources. Encourage them to seek diverse sources and to examine labor and globalization dynamics (the study notes the broader labor context around AI training and moderation).
Acknowledge language diversity and accessibility: The authors note multilingual discourses and the global nature of prompt conversations. In teaching, invite students to explore prompts in multiple languages or to design prompts that consider cross-cultural audiences. This aligns with a broader push toward inclusive, translingual technical communication.
Foster collaborative, reflexive methodology: The study emphasizes that researchers’ positionalities matter. In classroom settings, encourage group work that foregrounds different perspectives, and include student-research practices—e.g., a small project where learners document how collecting and analyzing prompts changes their own writing practices.
Ethics as a core thread: The research underlines ethical issues—privacy, the nuances of posting content from public spaces, data sharing, and the labor conditions behind AI systems. Build ethics check-ins into prompt-writing projects, and discuss what it means to publish or share prompts in open spaces versus keeping them private or locally useful.
Connect to real-world outcomes and societal stakes: The study’s discussion of labor issues (e.g., content moderation work in other countries) reminds us that technology is entangled with politics and economics. Encourage students to consider how their prompt-writing practices could affect workers, developers, and end-users in various contexts.
What this means beyond the classroom: real-world takeaways
Prompt writing is a social practice, not just a technical trick. Understanding its rhetorics helps you see who gains, who loses, and how language shapes outputs in business, media, education, and public discourse.
The culture around prompts includes micro-literacies, marketing narratives, and ethical considerations. Stay curious about who curates the resources you use, and how those resources encode assumptions about expertise, labor, and value.
A “wayfinding” mindset helps you stay flexible. Instead of chasing one perfect prompt, you learn to adapt prompts across contexts, audiences, and goals, while remaining critically aware of the limitations and biases in AI systems.
Critical literacy with AI means more than better results. It means asking questions like: Who is the audience? What are the potential biases in prompts? How might prompts reproduce inequities in language, labor, or access?
Practically, this translates into classroom and workplace activities: collaborative prompt analysis, diverse prompt databases, explicit ethics discussions, and iterative practice with feedback loops that foreground human judgment and cultural sensitivity.
Limitations and directions for future exploration
No study is perfect, and Gupta and Shivers-McNair are transparent about the evolving landscape of social media research and AI tools. Some key caveats to keep in mind:
Platform dynamics change: The dataset was collected when Twitter’s API access was open for research. Changes to platform access and data policies mean researchers—and teachers—may need new methods to study prompting discourses in the future.
Language and representation: The data show multilingual signals and global conversations, but the search terms were English, which may skew the sample. Future work could broaden to more languages and local contexts to better capture global prompting practices.
Depth vs. scale: The large dataset helped surface broad patterns, but the deep, qualitative work focused on 100 highly influential posts. Ongoing classroom projects can balance breadth and depth by combining large-scale data exploration with in-depth student analyses of prompts in chosen communities.
Ethics and sharing: The authors stress safety and aggregated reporting. In teaching, it’s important to model ethical data practices, respect privacy, and avoid sharing prompts that could reveal sensitive information.
A note on the scholarly lineage
The study sits at the intersection of digital writing, rhetoric, and AI literacy. It nods to a broader conversation about “writing with and for AI” and the push from professional organizations to treat prompt writing as a literacy practice. The authors explicitly connect their work to a tradition that values reflexivity, ethical awareness, and the recognition that literacies are dynamic, distributed, and embedded in social and political realities.
Key takeaways in plain language
Prompt writing is a broad, social practice that touches many areas of writing—from marketing and journalism to emails and coding. It’s not just about making ChatGPT spit out a better sentence; it’s about how people communicate in a world where AI is a writing partner.
There’s a growing ecosystem of micro-literacy resources (prompt databases, tips, tweetorials) that people share to help others prompt better. These resources are powerful but should be read critically and used thoughtfully.
Market rhetoric surrounds prompting: many posts treat prompts as a valuable skill or productivity booster. This has implications for teaching and ethics—who benefits, what’s being sold, and what assumptions about work are being promoted?
Prompts carry rhetorical characteristics: length, tone, templates, and even politeness cues matter. Prompt writing sits in a space between conversation and instruction, which has interesting implications for how we teach it.
Definitions of prompting vary: some see it as a usability improvement, others as a skill, or as a way of thinking about human-computer collaboration. Encouraging students to articulate their own definition helps them frame prompts for diverse contexts.
A wayfinding approach to teaching prompting can prepare students to navigate AI literacies across contexts. It invites collaboration, critical reflection, and practical experimentation with real-world prompts.
Teaching prompts well means acknowledging and addressing real-world issues like labor practices in AI, multilingual contexts, and the ethical use of shared resources.
If you’re an educator, a student, or just someone curious about how everyday conversations about ChatGPT shape our digital world, this study offers a thoughtful map. It doesn’t pretend there’s a single right way to prompt. Instead, it invites us to wander with intention, question our assumptions, and build a flexible, responsible set of prompting practices that can travel across genres, communities, and cultures.
Key Takeaways
Prompt writing is a dynamic social practice that influences a wide range of writing tasks, not just chatbot outputs.
Analyzing social media discourses about prompts reveals five themes: areas of communication impacted, micro-literacy resources, market rhetoric, rhetorical characteristics of prompts, and definitions of prompting.
A wayfinding, reflexive pedagogy helps teachers design activities that develop critical AI literacies, combining analysis of existing prompts with hands-on prompt design.
Micro-literacy resources (prompt databases, tips, tweetorials) are powerful but should be examined critically for accessibility, context, and ethics.
Market rhetoric around prompts can shape what’s valued in prompting and how resources are promoted; educators should teach students to assess value, equity, and sustainability.
Prompt writing is a rhetorical act with tone, audience, and context; educators can use this to teach prompts as craft, not just commands.
Definitions of prompting vary (usability, skill, talking to machines, thinking), so students should articulate their own working definitions for different contexts.
Ethical considerations—including platform changes, data privacy, and labor implications in AI—should be central to any prompted writing project.
Practical classroom workflow: discuss power and labor, analyze prompts with Ranade et al.’s formula, build a small prompt database, and run collaborative, iterative prompt-design projects.
If you want to elevate your prompt-writing game—whether for class, a community workshop, or your own projects—start with curiosity, stay critical, and treat prompts as living, evolving instruments that connect language, technology, and human values.
