Last Thursday, a Stanford sophomore submitted an essay about Virginia Woolf that her professor couldn't quite place. The prose shimmered with insight, the analysis felt fresh, yet something seemed off—not wrong, exactly, but uncanny, like hearing your own voice on a recording. When pressed, the student revealed her process: she'd fed her rough thoughts to Claude, discussed the output with her study group, revised with GPT-4, considered styles ranging from the New Yorker to the Atlantic, combined two with Gemini, then rewrote everything in her own words. "It is my work," she insisted! "I set all original directions, enhanced many of the arguments, improved it all and even deleted everything that didn't feel right or my style. I orchestrated and produced it all! Essentially, I used AI to think better together, like a word processor but working together with an enhanced brain for the 21st century and like everyone else is doing! Most importantly too, I learned a lot!"
This confusion—about authorship, authenticity, learning and the boundaries of human intelligence—would not surprise N. Katherine Hayles. From her perch at UCLA, the 82-year-old literary theorist and historian of science has spent four decades preparing us for exactly this moment. Long before ChatGPT began ghostwriting term papers, Hayles was mapping the hidden history of how humans and technology have always shaped each other's capacities for thought, learning and literacy. Her framework doesn't just explain our current AI predicament; it suggests we've been asking the wrong questions all along.
The concept Hayles calls "technogenesis"—the co-evolution of humans and technology—reveals why debates about AI replacing human intelligence miss the point. We don't simply use tools; we and our tools transform each other through reciprocal causality. The medieval scribe's mind was shaped by copying manuscripts just as surely as the manuscript preserved his thoughts. The telegraph operator developed what Hayles terms "telegraphic consciousness"—a staccato, compressed cognition that mirrored the medium. And now, after decades of emails, texts, and tweets, we've developed what she identifies as "hyper attention": the ability to process multiple information streams simultaneously, at the cost of the sustained focus that print culture prized.
This isn't decline—it's adaptation. When educators lament that students can't concentrate on long texts, they're measuring contemporary cognition against obsolete standards. Today's students navigate information environments that would overwhelm their grandparents: tracking multiple browser tabs while texting, absorbing podcasts while gaming, synthesizing TikToks into coherent cultural analysis. Hayles argues this represents not shortened attention spans but different attentional architectures, optimized for environments where information flows rather than sits.
The Alchemist's Apprenticeship
To understand how Hayles became our most prescient guide through the maze of machine meaning-making, consider her unlikely metamorphosis. In 1966, she worked as a chemist at Xerox Corporation, tracking molecular reactions with scientific precision. But even then, something in the chemical bonds whispered of larger patterns—how separate elements fuse into compounds that transcend their constituents, how consciousness itself might be such a compound, awaiting the right catalyst.
When she turned from chemistry to literature, she carried this atomic vision: meaning, too, emerges from collision and combination. This insight—that information never exists without a body, that meaning emerges from material interaction—would become the cornerstone of her contribution to AI literacy. While others debated whether machines could "think," Hayles asked different questions: How does thinking happen across networks? What new literacies emerge when cognition becomes collaborative?
Ghosts in the Machine
Between 1946 and 1953, in conference rooms thick with pipe smoke and possibility, the future was being written in feedback loops and flowcharts. The Macy Conferences on Cybernetics gathered minds that would reshape the century: Norbert Wiener mapping the mathematics of communication, Margaret Mead observing information patterns in human cultures, John von Neumann designing machines that might think. They were conducting, Hayles would later realize, a kind of séance—summoning spirits that still haunt our hard drives.
These ghosts whispered three dangerous delusions that Hayles spent decades exorcising. First: that information exists independently of its material substrate, floating free like a soul awaiting bodies. Second: that humans and computers are essentially similar information-processing systems, different only in degrees of complexity. Third: that consciousness can be uploaded, downloaded, or otherwise separated from the messy realities of embodiment. These cybernetic dreams, born in those smoke-filled rooms, still seduce Silicon Valley's fantasies of digital transcendence.
Against these illusions, Hayles proposes "cognitive assemblages"—networks where human and artificial intelligence collaborate without erasing their differences. Unlike the cybernetic vision of minds as computers, assemblage thinking recognizes that different forms of cognition excel at different tasks. Large language models process patterns across vast textual corpora but lack embodied understanding. Humans excel at contextual judgment but struggle with comprehensive recall. Together, they might achieve forms of intelligence impossible for either alone.
The Dance of Deep and Hyper Attention
Hayles' distinction between "deep" and "hyper" attention has become foundational for understanding our cognitive transformation. Deep attention—sustained, focused, singular—flourished in the age of print, in library carrels and scholar's studies. Hyper attention—rapid, switching, parallel—emerges from digital environments where information cascades rather than accumulates.
Neither is superior; they're different instruments in cognition's orchestra. The mismatch causes real suffering. A Berkeley junior describes feeling "broken" because she can't read novels the way her literature professor expects. But the same student created a brilliant final project analyzing how TikTok creators remix Victorian novels, tracking patterns across thousands of videos with a sophistication impossible in the deep-attention era. Hayles would argue she's not broken—she's adapted to a different information ecology.
Some educators are beginning to design what Hayles calls "assemblage pedagogies" that leverage both attentional modes. At Harvard's School of Education, teachers create "cognitive playlists"—lessons that alternate between deep-focus activities and hyper-attention tasks. Students use AI to scan hundreds of texts for patterns, then engage in sustained analysis of selected passages. The machines handle the hyper-attentional heavy lifting; humans provide the deep dive into meaning.
Material Semiotics: When Silicon Speaks
Hayles' approach to machine meaning-making draws on Charles Sanders Peirce's semiotics but with a material twist that changes everything. Traditional philosophy asks whether AI "understands" language—a binary question demanding yes or no. Hayles asks instead how meaning emerges from the physical processes of computation.
Consider what happens inside a large language model responding to a prompt. Electricity flows through circuits, activating weighted connections between artificial neurons. These weights encode patterns: "quantum" clusters with "physics," "uncertainty," "particles." The word "lovers" links to "passion," "distance," "connection." When GPT-4 generates a metaphor comparing quantum entanglement to "lovers who remain connected across impossible distances," something meaningful has occurred—not consciousness, perhaps, but genuine semiosis.
This is more radical than it sounds. Hayles isn't saying machines think like humans. She's saying meaning isn't exclusively human. Bacteria following chemical gradients engage in molecular meaning-making. Plants sending signals through mycorrhizal networks participate in forest-wide communication. Now artificial systems join this expanded community of meaning-makers—not conscious but cognitively significant.
The Embodiment Paradox
Every thought requires a body—this material fact punctures Silicon Valley's fantasies of uploaded consciousness. Hayles insists that information never floats free from its substrate. Your neurons firing as you read this sentence, the electricity powering the servers storing this text, the rare earth minerals in your device: all cognition depends on material resources with real environmental costs.
Current estimates suggest a single ChatGPT query consumes roughly as much electricity as lighting a LED bulb for an hour. Training large language models generates substantial carbon emissions—GPT-4's training has been compared to hundreds of transatlantic flights. As universities rush to integrate AI, Hayles asks uncomfortable questions: What are the carbon costs of thinking with machines? How do we balance cognitive enhancement against environmental destruction?
This ecological dimension transforms AI literacy. At the University of Toronto, students calculate the carbon footprint of their AI-assisted projects. One discovered her machine learning model for predicting crop yields consumed more energy than the farming techniques it aimed to optimize. "I thought I was solving climate change," she reflected. "Instead, I was accelerating it." Hayles would see this not as reason to abandon AI but to develop "ecological assemblage intelligence"—partnerships that account for their full material costs.
Posthuman Pedagogies in Practice
The traditional university—with its individual grades, solo examinations, and clearly bounded disciplines—assumes the liberal humanist subject Hayles argues never existed. Students always thought with libraries, collaborated with peers, extended their minds through notes and notebooks. AI simply makes visible what was always true: learning happens across networks, not within isolated minds.
Some institutions are redesigning around this insight. The London Interdisciplinary School (founded 2021) emphasizes interdisciplinary problem-solving that could align with what Hayles might call "assemblage thinking." Students tackle complex challenges—climate change, inequality, automation—that require orchestrating diverse forms of intelligence. This represents the kind of pedagogical evolution Hayles' framework suggests: learning to coordinate human creativity with machine analysis, recognizing when deep attention serves better than hyper attention, seeing ourselves as nodes in knowledge networks rather than containers of information.
The resistance comes not just from traditionalists but from students themselves. "I want to know what I know," explains Jordan Martinez, a philosophy major at Yale. "If I'm always thinking with machines, how do I know what's mine?" It's a profound question that Hayles addresses through her concept of "intermediation"—the recursive feedback loops through which humans and technologies reshape each other. What's "yours" was always mediated by language, culture, technology. The question isn't ownership but participation: How do you contribute to the cognitive assemblages you join?
The Ethics of Entanglement
Traditional ethics assumes autonomous agents making independent choices. But in cognitive assemblages, agency distributes across networks. When a medical AI recommends treatments based on biased training data, when educational software tracks and predicts student performance, who chooses? Hayles proposes "assemblage ethics" that traces responsibility through networks without reducing it to any single node.
Healthcare institutions demonstrate how this might work in practice. When hospitals use AI to predict which patients might miss appointments, Hayles' framework suggests mapping what she would call the "cognitive archaeology" of each prediction: What data trained the model? Which populations were overrepresented? How do human staff interpret predictions? By making the assemblage visible, they can address biases that would remain hidden in black-box systems.
This approach challenges romantic notions of individual genius that still dominate cultural discourse. When painter Marcus Chen collaborates with DALL-E to create "impossible landscapes," feeding the AI descriptions of dreams and painting its visual interpretations by hand, who is the artist? Chen insists the question misses the point: "It's like asking whether the painter or the brush creates the painting. We're creating something neither could imagine alone."
Critical AI Consciousness
The new literacies Hayles envisions go beyond technical skills. Just as print literacy meant more than decoding letters—it included understanding genres, evaluating sources, recognizing bias—AI literacy demands sophisticated capacities. Students must develop what Hayles calls "comparative media consciousness": recognizing how different substrates shape meaning. A Wikipedia article, a ChatGPT response, and a peer-reviewed paper carry different epistemological weights, not because one is "true" and others false, but because they emerge from different cognitive assemblages with different capabilities and constraints.
Sarah Kim, who teaches Critical AI Literacy at the University of Michigan, operationalizes these insights. Her students learn to recognize "algorithmic signatures"—telltale patterns revealing AI participation. They analyze how training data shapes responses, how prompts elicit different outputs, how human judgment remains essential for evaluating machine-generated content. Most importantly, they learn to see themselves as conductors orchestrating diverse intelligences rather than solo performers proving individual mastery.
The Planetary Perspective
As Hayles argues in her latest work, humanity faces challenges—climate change, pandemics, inequality—that exceed any individual intelligence, human or artificial. Solutions require what she terms "planetary cognition": coordination across human communities, AI systems, and the living networks that sustain civilization. The student learning to think with machines today might tomorrow orchestrate the cognitive assemblages addressing existential threats.
Yet dangers lurk in this posthuman paradise. When cognition distributes across networks, who bears responsibility for outcomes? If a medical AI assists in misdiagnosis, whom do we blame? More troubling: as we adapt to thinking with machines, what capacities might we lose? The GPS example haunts—London taxi drivers who navigate by memory have larger hippocampi than those relying on apps. What cognitive muscles atrophy as we outsource functions to AI?
Hayles acknowledges these risks but argues we've always made such trades. Writing weakened memory; calculators diminished mental arithmetic. The question isn't whether to change but how to change thoughtfully, maintaining what matters while gaining new capacities.
Coda: Learning the Steps
This brings us back to that Stanford sophomore, defending her AI-assisted essay about Virginia Woolf. In Hayles' framework, she's not a cheater but a pioneer, fumbling toward new forms of intellectual partnership. The question isn't whether her work is authentic but whether it's thoughtful—whether she's orchestrated the assemblage with wisdom, creativity, and critical judgment.
The posthuman classroom Hayles envisions doesn't arrive through institutional decree but through thousands of such experiments. Each student learning to think with machines, each teacher designing assemblage pedagogies, each institution reimagining assessment—together, they're creating the educational future Hayles spent four decades preparing us to inhabit.
We are not becoming posthuman; we are recognizing we always were. The dance between minds and machines isn't beginning; we're just finally learning to hear the music. And in that recognition lies both loss and liberation—the end of illusions about autonomous intelligence, the beginning of possibilities for collaborative cognition that might, just might, prove adequate to the challenges that await.
Brief Annotated Bibliography
Primary Works by N. Katherine Hayles
Bacteria to AI: Human Futures with our Nonhuman Symbionts (2025) a summary and continuation of themes based on Hayles' intellectual trajectory continuing her themes into the 21st century and AI (summarized concepts in the article above).
How We Became Posthuman: Virtual Bodies in Cybernetics, Literature, and Informatics (1999) The foundational text that traces the historical development of the posthuman through the Macy Conferences on Cybernetics. Introduces the "three stories" of how information lost its body, the cultural construction of the cyborg, and the emergence of the posthuman subject. Essential for understanding her critique of disembodied information.
Writing Machines (2002) Explores the materiality of texts and introduces the concept of "technotexts"—works that foreground their own material properties. Demonstrates how the physical form of texts shapes meaning, laying groundwork for understanding AI as material rather than abstract intelligence.
My Mother Was a Computer: Digital Subjects and Literary Texts (2005) Develops the theory of "intermediation" between human and digital systems. Contains crucial early thinking about distributed cognition and the recursive feedback loops between humans and technologies that would later inform her approach to AI.
How We Think: Digital Media and Contemporary Technogenesis (2012) Introduces the distinction between "deep attention" and "hyper attention" that has become foundational for understanding cognitive changes in digital environments. The concept of "technogenesis"—the co-evolution of humans and technology—is fully developed here.
Unthought: The Power of the Cognitive Nonconscious (2017) Expands the definition of cognition beyond consciousness to include nonconscious processes in both humans and machines. Prepares theoretical ground for understanding AI as a form of nonconscious cognition. Introduces "cognitive assemblages" as networks of human and technical actors.
Key Theoretical Influences
Charles Sanders Peirce, "Logic as Semiotic" (1897-1910) Provides the triadic theory of signs (icon, index, symbol) that Hayles adapts for understanding machine meaning-making. His material approach to semiotics informs her analysis of how AI systems create meaning through physical processes.
Gilles Deleuze and Félix Guattari, A Thousand Plateaus (1980) Source for assemblage theory, though Hayles emphasizes material cognition over their focus on deterritorialization. Her "cognitive assemblages" adapt their concept for analyzing human-machine collaborative intelligence.
Secondary Sources on Posthuman Pedagogy
Siân Bayne, "Posthumanism and Research in Digital Education" (2018) Applies posthuman theory to educational technology, drawing heavily on Hayles' framework to analyze how digital tools reshape learning. Provides practical examples of posthuman pedagogical approaches.
Casey Boyle, Rhetoric as a Posthuman Practice (2018) Examines writing instruction through a posthuman lens, using Hayles' concepts to argue for teaching writing as assemblage practice rather than individual expression.
Note on Examples
The institutional examples and specific programs mentioned throughout this essay (including the opening Stanford anecdote) are illustrative scenarios that demonstrate how Hayles' theoretical framework might be applied to contemporary educational challenges. While some institutions are indeed exploring posthuman approaches to AI literacy, the specific cases presented here serve as thought experiments rather than documented implementations. They show how her concepts could transform pedagogical practice as educators grapple with AI integration. Readers interested in specific programs should verify current offerings at individual institutions.
For Further Reading
Hayles' UCLA faculty page and the Electronic Literature Organization archive contain additional resources. Her perspectives on AI and education appear in various academic journals and literary reviews. The N. Katherine Hayles Collection at Duke University's David M. Rubenstein Rare Book & Manuscript Library houses her papers and correspondence.
#KatherineHayles, #AILiteracy #Posthumanism #AI #LLMs