Writing: An Ancient Craft in a Brave New World
"Señor? Señor? Can you tell me where we're headin', Lincoln County Road or Armageddon? Seems like I've been down this road before" (Bob Dylan, Señor).
Writing has always been a tango between thought and expression, but in today’s AI-driven world, the tools we use to write—and think about writing—are evolving at an unprecedented pace. As we navigate this shift, it’s clear that the way we talk about writing and reading needs to evolve, too. That’s why, in this piece, I’ve chosen to use scientific terms like syntax and readability formulas instead of simpler alternatives.
My goal isn’t to alienate anyone but rather to encourage us all to expand our shared vocabulary for discussing these concepts. After all, if we’re going to embrace the full potential of AI tools in writing, we need language that matches their complexity and the complexity of literacy processes and practices.
I get it—some of these terms might feel a bit academic or unfamiliar. I read Substack posts, too, and some of them are so dense that I have to slow down and draw on resources to comprehend them. But here’s my suggestion: Keep an AI tool open as you read. Tools like ChatGPT or Claude can help you quickly find simpler explanations or synonyms for any term that feels too dense. That’s not cheating. That’s good common sense.
For example, if "syntax" trips you up, ask your AI assistant for a plain-English definition. This isn’t just about understanding this article—it’s about equipping yourself with the tools to engage with writing and reading in a more nuanced way. Syntax isn’t a peripheral, ten-dollar word; it refers to a fundamental knowledge requirement if you are to fully grasp issues like the recent outlawing of the three-cueing system in early reading instruction across America.
So, as you dive into this exploration of how writing tools are reshaping our cognitive and creative processes, I invite you to think of this as an opportunity to upgrade your own writing vocabulary toolkit if you need to—not just with new ideas but with the words to express them. Let’s embrace the challenge together and roll up our sleeves.
The Evolution and Integration of Writing Tools
The tools we use for writing in today's digital environment extend far beyond physical devices like keyboards and screens. These physical implements remain essential, and they make writing more efficient and text more malleable than tools from the Stone Age, but a new, even more influential category of hybrid tools has emerged over the past fifteen years that bridges the physical and cognitive domains, ranging from spellcheck to autoprediction. Software that analyzes text readability level (first grade, eighth grade, college) can rate levels of vocabulary difficulty and syntactic challenge to scaffold a writer's stylistic decisions regarding linguistic complexity during the writing process. Readability software produces scores to describe text difficulty for readers: 3.5 means a reader in the middle of third grade could be expected to comprehend the text.
Software applications like readability tools operate at the intersection of technology and human thought processes. Such hybrid tools do more than facilitate the transfer of ideas to and from text more efficiently; they shape how writers compose and how readers engage with written content. Judging whether a simpler synonym can accurately replace a more complex term or whether a subordinate clause clarifies or complicates the reading process goes well beyond simple efficiency into effectiveness. Understanding these hybrid tools and their impact requires us to expand and sharpen our traditional definitions of writing implements and examine how they influence the entire writing process.
Beneath external tools like keyboards and software lies a more fundamental set of tools shared by readers and writers. These cognitive (thought) mediators work invisibly, influencing how writers develop ideas and how readers engage with completed texts. The traditional approach to teaching outlining, for example, illustrates how physical constraints can overshadow cognitive tools in writing instruction. When teachers require students to create and formalize outlines with rigid hierarchical structures before drafting and expect writers to follow their outline, they transform what should be a flexible thinking tool into a prescriptive physical device, a blueprint, that can inhibit the writing process. Writing becomes a process of recording pre-formed thoughts, not a complex cognitive activity that transforms both the writer's understanding and the reader's experience.
Consider how this manifests in a typical classroom in bygone years and even in current classrooms: Students produce a detailed outline with Roman numerals, upper case and lower case letters, and a staircase of indentations before they begin composing their draft. This approach reveals a simplified understanding of the cognitive role of outlining. Rather than allowing students to use organizational frameworks as malleable tools for cognitive exploration and developing ideas, it presents the outline as a fixed job description to govern the composing process.
This misapplication has significant consequences for teaching and learning to write. Students who create outlines to satisfy format requirements rather than to support their thinking process may struggle to modify their initial organizational structure as their ideas evolve during writing, feeling constrained by their preliminary outline rather than empowered by it.
This constraint can be amplified when students must get approval, sometimes a grade, on their outline from their teacher before they are allowed to begin drafting. The outline isn’t best viewed as a finished product, but as a tool. The result of seeing an outline as a discrete step which must be completed before progressing to the next step is to disconnect thought from the development of ideas that scholars know occurs during composition.
A more effective approach would be to recognize outlining as a cognitive tool that evolves alongside the writer's understanding. Thinking of an outline as the roughest of rough drafts could be useful. This way of thinking means teaching students to view organization, a critical aspect of a finished text, as up in the air until the final bell rings.
Computers make it easy to move whole sections, paragraphs, and sentences. Expecting their outline to develop and change as their ideas mature, writers have the tools and the permission they need to monitor the emerging macrostructure of the text that will eventually stabilize. Such an approach acknowledges that effective organization emerges from the interaction between writer and text, rather than being imposed before writing begins.
Writing Tools in Professional and Academic Contexts
Assuming that physical, hybrid, and cognitive mediators shape the writing process, educational researchers and teachers might collaborate to build a new theoretical framework that explains the complex relationships between writers, readers, and the tools they employ. Theoretical frameworks are not meant to guide pedagogy, but to explain how a phenomenon like “writing” works in such a way that professionals understand one another.
This new framework must account for how writing tools serve multiple functions simultaneously: Supporting writers as they develop their ideas while also facilitating readers' comprehension of the finished text once stabilized and durable. Understanding these dual roles is essential for recognizing how writing tools mediate the entire literacy process, from initial conception through final understanding.
Writing tools serve as mediators between writers and their emerging text and between readers and their comprehended text within specific task environments. Some tools are closely integrated in task environments. Readability software like Grammarly, for example, has become important in the work of technical writers who aim for the least complicated, most transparent version of complex ideas and for readers who can take advantage of transparency and read more efficiently and effectively in areas where they don’t have expertise or want just the facts.
The thesis statement in academic writing, as another example, has a dual-mediating role as a literacy tool: These statements mediate between (1) writers and their emerging text, helping them arrive at a commitment to a central idea for the text to become their intention and purpose, and (2) readers and their emerging mental model of the text, who rely on it as a hub for their reconstruction of meaning, all within specific writing task environments. Because the thesis statement plays such a dominant role in high school writing instruction today, what it is, how it works, why it works, and how it is taught most effectively could be extremely important in interaction with digital tools.
Creative nonfiction writers in the field of journalism employ several distinct categories of writing tools that shape both their cognitive processes and final texts and can be supported by AI tools. Narrative structuring tools help writers organize lived experiences into compelling stories while maintaining factual accuracy, including techniques like scene-setting, character development, and narrative arc construction.
Research and memory tools assist creative nonfiction writers in gathering and processing source material, including the journalist's own emotions and sensations during an event. Going on anAI “previsit” before taking a physical trip to experience Rodin’s work could be helpful in priming the pump of ideation even before the visit to do research for a feature article. YouTubes and documentary films could help, but they don’t take your questions, neither do they respond to them.
Voice development tools help writers craft their distinctive authorial presence as a participant and an observer of history. These cognitive mediators include strategies for modulating tone, managing the distance between writer and subject matter, and achieving a balance of personal reflection with objective observation. Ironically, the creative nonfiction writer presents a text intended to be read as a conveyor of spontaneous and highly personal meanings while actually being the product of careful crafting and strategic construction. These writers must constantly navigate between their role as the eyes and ears of the reader, their own eyes and ears as a reader surrogate, and the recorder of fact.
As digital technologies evolve, Large Language Models (LLMs) represent a totally new category of hybrid tools for writing that further blurs the distinction between physical and cognitive mediators. AI tools demonstrate sophisticated capabilities in analyzing text accessibility and audience appropriateness. When prompted to evaluate complex text from a specific reader's perspective, such as a seventh-grade student, these tools can provide detailed recommendations for improving comprehension.
For instance, AI might suggest replacing abstract terms like "cognitive domains" with more accessible phrases like "how our brain thinks and works," or recommend breaking complex sentences into simpler structures. This capability extends beyond traditional readability formulas by offering context-sensitive suggestions that consider both vocabulary and sentence complexity while maintaining the text's essential meaning. Such interactive analysis exemplifies how AI tools function as collaborative partners in the writing process, helping writers bridge the gap between their expert understanding and their readers' needs.
Consider the application of custom-trained LLMs in literary scholarship: researchers can now develop specialized models trained on comprehensive literary corpora, such as Shakespeare's complete works, to conduct sophisticated thematic analyses across entire bodies of work. This capability transforms traditional literary analysis by enabling scholars to identify patterns and connections that might otherwise remain obscure through conventional research methods. For example, a researcher might use a customer-trained Shakespeare bot to examine the characteristics of all of the fools in Shakespeare’s play.
Larger commercial LLMs like ChatGPT and Claude serve as versatile writing collaborators, offering capabilities that extend beyond traditional writing tools. They can function as sounding boards for ideas, help writers clarify their thinking, and provide alternative perspectives on developing texts. However, their role as mediators is more complex than previous writing technologies, for they actively participate in the cognitive aspects of writing while simultaneously serving as tools for simulated text production.
Search-augmented AI platforms like Perplexity introduce yet another dimension to writing tools by combining real-time information retrieval with natural language processing. These systems help writers bridge the gap between research and composition, offering immediate access to current information while maintaining the interactive qualities of AI assistants. This integration of search and dialogue creates a new kind of mediation between writers and their source materials.
The emergence of these AI writing tools raises important questions about the nature of authorship and the writing process itself. Writers must now develop new cognitive strategies for collaborating effectively with AI systems while maintaining their original intention and purpose for writing and ensuring the integrity of their work, a different criterion than the authenticity of their work. As these tools continue to evolve and reshape writing practices, we must examine their impact through systematic research to understand their full potential and limitations.
Research Directions for AI-Enhanced Writing
Thesis Development and AI Assistance
Research into thesis development with AI assistance represents a particularly promising area of inquiry. Investigators might examine how writers interact with AI systems to develop and refine thesis statements through multiple iterations, comparing the quality and effectiveness of AI-assisted thesis development with traditional methods. Such studies could track how thesis statements evolve through AI-supported drafting and how readers use and respond to these statements. This research could reveal whether AI assistance leads to clearer, more focused thesis statements that better guide both writers and readers through complex arguments.
AI's Role in Outlining and Organization
The role of AI in transforming outlining and organizational processes presents another rich area for investigation. Researchers could explore how writers use AI-enhanced outlining practices that allow for more flexible, dynamic organization of ideas. For example, I have a hypothesis that writers who informally map ideas for a writing project on their own before consulting a bot produce more effective thesis statements when using AI than do students who bypass this initial human work and start outlining interactively with a bot. This hypothesis is testable.
Studies might compare the organizational patterns that emerge in final texts when writers use AI assistance after training on prompting strategies versus traditional outlining methods under different circumstances (e.g., a formal outline that is graded and must be used versus an outline that students use as a malleable and flexible tool in both conditions uninformed by AI). There would be three groups in the study: Two control groups creating traditional outlines, one group formal outlines, the other group freeform outlines, and one treatment group using AI.
Examining how writers maintain their own organizational logic under compliance with teacher restrictions vs. the liberty to revise their outline over the course of the task could be important baseline information that may interact with results when access to AI takes place in subsequent research. This research could illuminate how AI tools affect writers' willingness to restructure their documents during development.
Comparing AI-Assisted and Independent Draft Production
How do thought processes used to produce rough drafts work as springboards to a product differ when writers compose rough drafts with AI assistance versus working independently, specifically examining variables such as ideation fluency, organizational coherence, alignment with the writer’s intentions, and writers' sense of ownership over their texts? Do readers detect differences between final texts produced under each condition?
This research question could be investigated through a controlled study where participants complete two comparable writing tasks—one with AI assistance and one without—while researchers gather data through think-aloud protocols, screen recordings, and textual analysis to measure factors including time spent drafting, patterns of revision during the drafting process, and key characteristics of the resulting drafts as well as reader response data.
AI Tools in the Revision Process
Revision processes with AI support offer particularly compelling research opportunities. First, distinguishing between local revisions and global revisions could produce different findings for traditional methods of instruction as well as AI-integrated practices. Studies could track how writers use AI tools to identify problem areas needing clarification or development in early drafts and examine the patterns of activity that emerge in revision cycles when writers can query AI about specific aspects of their drafts versus when they cannot.
Designing scales for measuring individual writer’s skill as AI prompters would be important. Research has already shown wide discrepancies in value of bot exchanges depending on the expertise of the user. Researchers might compare final drafts between high-skilled users of AI for selective revision tasks versus those who have just an emerging level of AI skill. What differences in revision qualities can be discerned when writers target particular points in a text with a specific question versus writers who ask for holistic feedback? Are there more effective and less effective strategies for approaching revision? What are the risks of bot takeover during revision?
Reader Response and Text Effectiveness
The impact of AI on writers’ audience awareness presents complex questions about how writers use these tools to anticipate and respond to reader needs. Research could examine how writers use AI to test different versions of their text against simulated reader responses and how these simulation responses compare to human reader feedback. Such studies might reveal how AI-assisted audience analysis affects writers' revision decisions and ultimate level of success in reaching their intended readers.
Studies would then extend to examining reader response of the finished text with a diverse group of readers evaluating the products across multiple dimensions: clarity of argument and organization, depth and sophistication of ideas, aesthetic appeal of the writing style, perceived credibility of the author, and overall effectiveness. Readers would engage in both quantitative assessment using standardized rubrics and qualitative response through think-aloud readings and semi-structured interviews, allowing researchers to understand how AI assistance during revision ultimately affects the reader's experience of the text-product.
AI and Reading Complex Texts
In the domain of reading comprehension of informative texts as well as reader responses to literary texts, researchers could investigate how AI affects students' engagement with different kinds of complex texts. Studies might examine the ways in which readers might use AI to generate visible images of descriptive passages in literature or detailed specifications of parts of a process or mechanism in science. Eperiments could be designed to discern whether such a use enhances readers’ inclination and capacity to visualize during reading after AI experiences.
Studies might examine differences in comprehension when students use AI for targeted text queries versus traditional annotation methods, or how AI-assisted close reading affects students' development of independent analytical skills. Of course, close reading must be defined in testable ways; its current widespread use tends to be idiosyncratic and localized.
To understand AI's impact on literary envisionment building, researchers might examine how students engage with complex modernist texts such as Virginia Woolf's Mrs. Dalloway. Such studies would assess how AI-assisted reading influences students' analytical capabilities, measuring both their technical understanding of modernist techniques and their capacity for meaningful personal response to the text. This research would help determine whether AI tools enhance or potentially inhibit students' development of independent literary analysis skills.
Changing Fear and Trembling into Actionable Research
The evolution of writing tools—from physical implements to hybrid and cognitive mediators—has fundamentally reshaped how we write, think, and read. As AI technologies like Large Language Models continue to blur the lines between tools and collaborators, we find ourselves at a pivotal moment in the history of literacy. These tools are no longer just aids; they are active in the creative process, challenging our traditional notions of authorship, organization, and even comprehension.
But with this transformation comes responsibility. Writers, educators, and researchers must embrace these changes thoughtfully, developing new strategies to collaborate and co-think with AI while preserving the integrity of human creativity and critical thinking. Whether it’s reimagining outlines as flexible cognitive tools even more helpful (or harmful) with AI or leveraging AI to refine readability and audience engagement, we must approach these innovations as opportunities to expand—not replace—our own capabilities.
I’m not talking about adapting to new tools; I’m talking about evolving the way we think about writing pedagogy itself. How can we use these technologies to enrich our understanding of text production in the future? How can these technologies help us understand ways to use new affordances to evoke greater depth and clarity in reading? These are questions worth exploring together.
So let’s get to it. Whether you’re a writer experimenting with AI or an educator rethinking how you teach composition or a parent trying to advise your own children on appropriate uses, this is your invitation to join the conversation. I’ve left a leave a comment button below. The future of writing is being shaped right now—and we all have a role to play in defining it.