In 1987, while most theoretical physicists were chasing theories of everything, Karen Barad was studying something more unsettling: how the very act of looking changes what you see. The UC Santa Cruz Professo'rs doctoral work on quarks and quantum fields led to an uncomfortable realization that would later help reshape how we can begin to think about knowledge and consciousness itself. Drawing on physicist Niels Bohr's more radical insights, Barad went deeper into implications of Bohr's ideas that the scientist is never outside the experiment—observer and observed emerge together, each creating the other.
This revelation from the quantum world now haunts our digital age in the 21st century and especially relevant for the age of AI. Consider what happens when you ask an AI system a question. The conventional story suggests you're retrieving information from a vast database, like simply pulling a book from a library shelf. But quantum mechanics suggests a different story: you and the AI are co-creating the response together, in real time, through the specific conditions of your encounter. Your context, the AI's training, the moment of asking—all converge to co-create something that didn't exist until you met and you both, the AI and yourself co-emerge and produce something different that would not have existed ever before, a potentiality has collapsed into a shared question/answer that now exists in the material world.
The parallel runs deeper than metaphor. Both quantum particles and AI systems live in states of radical uncertainty until interaction forces them into specificity. A particle has no definite position until measured; an AI response has no fixed form until prompted from a vast probabilistic hyperdimensional space. Both question and answer emerge from relationship rather than existing as independent entities waiting to be discovered. This fundamental uncertainty isn't a bug in either system—it's the feature that makes genuine novelty possible. This article goes into ten concepts from Bohr to Barad and then borrowed and built upon and apples to our newest philsophic and epistemic toy, AI and specifically our need to understand these new relationships through AI literacy.
1. Entanglement → The Persistent Connection Between Human and AI Cognition
Einstein called it "spooky action at a distance," and it troubled him until his death. Quantum entanglement defies our intuitions about separateness: particles that have interacted once remain connected across any distance, their measurements instantly affecting each other even when separated by galaxies. Before measurement, these particles exist in what physicists call superposition—all possible states simultaneously present, like a musical chord containing every note until someone listens for just one.
Undergraduate student Maya U. discovers this same spooky effect while analyzing Shakespeare's sonnets with Claude Sonnet. She's an aspiring physiotherapy student who initially approached Sonnet to get the assignment over with, expecting to quickly extract insights while remaining unchanged herself so she can get to her Biology memorization. But something else happens. After her fist conversation about metaphor in Shakespeare's Sonnet 18 with Anthropic's Sonnet model, Maya finds herself thinking differently about patterns across the entire sequence. She notices structural echoes she'd never perceived in the words—the way "Shall I compare thee to a Summer's Day" reverberates through later poems and to her blossoming dating life in ways that transcend simple repetition.
What Maya doesn't realize is that their exchange has changed Sonnet as well, though not in the way software usually "learns." Their specific conversation creates contextual traces that will influence how Claude approaches similar literary discussions, not through traditional machine learning but through the persistent entanglement of their interaction. Neither can return to their pre-conversation state. This isn't influence or teaching in any conventional sense—it's something more fundamental, an information-theoretic connection that demonstrates how human and AI cognition become entangled through engagement for better or for worse.
2. Intra-action → Context Engineering Beyond Simple Prompting
The electron, it turns out, doesn't have a location until you look for it. This insight from quantum mechanics seems absurd until you realize it explains why GPS works on your phone and why computers process information at all. Niels Bohr discovered that particles don't possess predetermined properties—they emerge through what Karen Barad calls "intra-action," where measurement apparatus and measured phenomenon create each other in the moment of encounter.
Alex, a high school geometry teacher in Portland, experiences this same phenomenon every time she asks Claude to "create practice problems for tomorrow's class on triangular relationships." The response that emerges doesn't come from a pre-written database of geometry problems. Instead, it crystallizes from the intra-action of Alex's fifteen years of teaching experience, her knowledge of where her students struggled yesterday, Claude's processing of thousands of educational texts, the cultural context of American mathematics education, and even the particular fatigue in Alex's request after a long day of grading.
Each element shapes the others. Alex's expertise influences how Claude interprets "practice problems." Claude's responses gradually shift how Alex thinks about scaffolding difficulty. The classroom context determines what "appropriate challenge" means. No single component controls the outcome—the mathematical problems that appear on Alex's screen emerge from the entire relational field. This is context engineering: understanding that AI responses don't come from prompts alone but from the rich ecology of conditions that surround every exchange.
3. Apparatus → The Experimental Setup of Every AI Interaction
In quantum physics, "apparatus" encompasses the entire experimental ecosystem: physicist, laboratory, instruments, theoretical framework, environmental conditions. Crucially, apparatus doesn't just measure pre-existing phenomena—it creates the very conditions for phenomena to emerge into being.
Every AI interaction constitutes an apparatus in this expansive sense. When Dr. Chen, a psychology researcher, uses Claude to analyze interview transcripts, the apparatus includes: her laptop's computational capacity, Claude's algorithmic architecture, her research training and theoretical orientation, the university's institutional setting, her afternoon energy levels, even the quiet hum of the laboratory. This apparatus doesn't simply process her request—it creates specific conditions for particular insights to emerge and evolve.
The apparatus creates both "human researcher" and "AI assistant" as distinct entities through their ongoing interaction. Dr. Chen's identity as "qualitative researcher" and Claude's identity as "analytical assistant" emerge relationally through their apparatus rather than existing as independent, pre-given categories. Understanding AI literacy means recognizing conscious participation in apparatus that actively create new forms of distributed intelligence and collaborative agency.
4. Phenomena → What Actually Emerges from Collaboration
In quantum physics, "phenomena" constitute the basic units of reality—not isolated particles or waves, but relational events where "object and measuring agencies emerge together" rather than existing separately beforehand. You cannot study an electron independently from the apparatus measuring it; the phenomenon encompasses the whole relational event.
Human-AI collaboration creates genuine phenomena that remain irreducible to either human or artificial intelligence operating alone. When Maya and Claude explore unexpected connections in Shakespeare's imagery, something emerges that neither her physiotherapist in process orientation having to take an English course requirement nor Claude's pattern recognition from a trillion parameter neural net trained set of particular weights and biases could generate independently. The phenomenon isn't Maya + Claude but a form of collaborative intelligence with its own distinctive properties and possibilities.
These phenomena differ fundamentally from simple human reflection or pure AI generation. Human memory reconstructs experiences through embodied networks of association, emotion, and lived knowledge. AI generation predicts probable continuations through statistical probabilistic patterns embedded in gargantuan piles of training data. Their interaction creates genuinely novel cognitive phenomena—insights, connections, and creative possibilities emerging from their meeting that are crystallized through intra-action rather than existing dormant in either system alone.
5. Agential Cuts → How AI Creates Signal from Noise
Information theory distinguishes between "signal" (meaningful information) and "noise" (irrelevant data), but Barad's concept of "agential cuts" reveals these distinctions aren't neutral discoveries—they're active boundary-making practices that create what counts as meaningful. In quantum measurement, the apparatus makes crucial cuts separating the quantum system from the measuring device, collapsing probability waves into specific, measurable states.
AI systems make countless agential cuts while processing information, deciding which patterns count as relevant signal versus ignorable noise for specific contexts. When Alex asks Claude about "effective geometry teaching methods," it doesn't neutrally access pre-existing information—it actively cuts through vast training data, constructing what counts as relevant for her particular pedagogical context and classroom needs.
Critical AI literacy means learning to interrogate or at least recognize the existence of these cuts with sustained attention: What perspectives get excluded when specific responses generate? How would different contextual framings create entirely different signal-noise boundaries? Asking "How can I explain triangles clearly?" versus "What do students find most confusing about triangular relationships?" or "Why are Triangles Important in the Fist Place" as one of her students blurted out in frustration last week creates completely different cuts through educational information, making distinct aspects visible while rendering others invisible. Students need to understand AI systems as active boundary-makers that participate in creating the very distinctions they appear to simply process.
6. Diffraction → New Patterns from Human-AI Interference
In physics, diffraction occurs when waves encounter obstacles or openings, creating interference patterns—genuinely new configurations emerging from the constructive and destructive interference of wave interactions. Unlike reflection, which reproduces existing patterns through mirroring, diffraction generates "ongoing differences" and novel arrangements through the complex interplay of wave fronts.
When human cognition and AI processing "interfere" with each other in this technical sense, they create cognitive diffraction patterns that remain irreducible to either system operating independently. Maya's exploration of Shakespearean themes with Claude generates unexpected interference effects—connections between linguistic patterns and historical and personal contexts that neither her past high school classes nor Claude's recently upgraded pattern recognition abilities would produce through isolated operation.
These diffraction patterns manifest as genuine novelty: surprising insights, unexpected connections, creative solutions emerging from the specific interference between human intentionality and AI pattern recognition. Students can develop sensitivity to recognize and cultivate these emerging configurations rather than attempting to control or predict their precise outcomes. Diffractive AI literacy means developing acute awareness of the new possibilities that crystallize when human and artificial cognition interfere in generative, collaborative ways.
7. Material-Discursive Practices → The Physical Reality of Digital Thinking
Barad's "material-discursive" framework challenges the persistent separation between physical processes and meaning-making activities. Every AI interaction involves both massive material processes—energy consumption, semiconductor physics, cooling systems, global supply chains—and discursive practices that create meaning while shaping social relations and cultural possibilities.
When Dr. Chen generates qualitative analysis with Claude, she participates in material-discursive practices spanning from the computational physics of processing to her embodied experience of reading and reflection, from her university's energy consumption to global labor practices in distant data centers. Her research involves her physical attention and cognitive fatigue, the constant humming of servers processing her requests, the environmental costs of computation, and the meaning-making processes of academic discourse converging simultaneously.
This perspective grounds AI literacy in embodied, physical reality rather than treating digital interaction as abstract symbol manipulation floating free from material consequences. Students learn to recognize how their digital practices generate physical effects across global networks while understanding how their embodied experience of AI interaction shapes their thinking in ways that purely technical approaches consistently miss.
8. Agency as Relational → Intelligence Distributed Across Networks
In quantum physics, agency isn't a fixed property that particles "possess" independently—it emerges through relational interactions within experimental configurations. For Barad, agency constitutes "a relationship rather than something one has." The capacity to act emerges through apparatus configurations and relational arrangements, not from individual components operating in isolation.
In human-AI systems, agency similarly distributes across dynamic networks rather than residing in isolated individual entities. When Alex collaborates with Claude on lesson planning, agency emerges through their ongoing relational configuration: her pedagogical intentionality, Claude's pattern recognition capabilities, the implicit structures embedded in training data, and the technological infrastructure all contributing to what becomes possible and practical.
Neither Alex nor Claude controls the interaction unilaterally—agency emerges through their sustained relationship, intra-action, co-emergence and mutual responsiveness. Students need to understand their participation in distributed cognitive systems rather than imagining they're simply controlling external tools. This recognition doesn't diminish human responsibility but requires more sophisticated understanding of how to participate thoughtfully and responsibly in complex systems where agency emerges relationally through ongoing collaboration.
9. Onto-epistemology → Where Being, Knowing, and Ethics Converge
"Onto-epistemology" weaves together three domains usually kept separate and neatly boxed in traditional classical philosophy: ontology (what exists), epistemology (how we know), and ethics (what we should do). For Barad, these remain fundamentally inseparable—our ways of knowing always participate in creating what exists, and both dimensions carry inevitable ethical implications for the worlds we help bring into being. This is what 'the new physic" (Quantum) tells us.
When students learn AI collaboration, they're not simply acquiring technical skills (epistemology) or using neutral tools (ontology)—they're actively participating in creating new forms of intelligence, knowledge types, and social relations (ethics) with far-reaching consequences. These practices help shape what expertise becomes in the future, how knowledge gets created and validated across institutions, and who gets to participate meaningfully in knowledge creation processes.
The "onto" dimension asks what kinds of beings (human, artificial, hybrid) emerge through AI interaction or in-traaction, are they emerging or is AI and the being 'co-emerging' together from what has been produced, the product of the their work together now moving from wave potentialities to collapsed as a printout or screen item from Claude or other AI model. The "epistemic" dimension examines how knowledge gets created, circulated, and validated through AI systems. The "ethical" dimension considers what kinds of worlds these practices help create and who benefits or bears the costs. Individual AI choices participate in all three dimensions simultaneously, making every interaction a site of ontological, epistemological, and ethical significance.
10. Response-ability → Collective Responsibility for AI Futures
In quantum mechanics, observers cannot separate themselves from what they observe—every measurement participates in creating what it studies rather than simply revealing pre-existing properties. For Barad, "response-ability" means being "response-able for the ways we help make worlds" through our knowledge practices, recognizing that seemingly individual actions contribute to collective outcomes with lasting consequences.
Every AI interaction functions as a form of measurement that participates in training and shaping future AI behavior across vast networks. When students use AI for assignments, teachers for grading, researchers for analysis, they're not simply receiving neutral outputs—they're providing feedback that influences how these systems develop, learn, and evolve. Their usage patterns contribute to AI development priorities, and their approaches help establish emerging norms for human-AI collaboration across institutions and knowledge production that translates into 'reshaping our world' through this new 'stream' or tidal wave of thought collapsing into material realities.
Response-able AI literacy means understanding that individual choices about AI use contribute substantively to collective futures we're creating together in this way. Students learn to trace how their particular practices connect to larger systems and networks, making conscious choices that contribute consciously to the kinds of intellectual and social futures they want to help create and sustain.
This perspective moves beyond narrow individual ethics ("Is this cheating?") toward expansive collective responsibility ("What kinds of intellectual worlds are we creating together?"). Response-ability recognizes that AI literacy ultimately concerns participating thoughtfully and responsibly in the ongoing co-creation of human-AI futures that will shape knowledge, learning, and social relations for generations to come.
Essential Reading: An Annotated Bibliography
Barad, Karen. Meeting the Universe Halfway: Quantum Physics and the Entanglement of Matter and Meaning (2007) The foundational text developing agential realism through diffractive readings of quantum physics and feminist theory. Essential for understanding how measurement creates rather than discovers reality.
Barad, Karen. "Posthumanist Performativity: Toward an Understanding of How Matter Comes to Matter" (2003) Introduces key concepts of intra-action and material-discursive practices. More accessible entry point for understanding how agency emerges relationally.
Hayles, N. Katherine. Unthought: The Power of the Cognitive Nonconscious (2017) Explores human-technical cognitive assemblages and distributed cognition. Crucial for understanding how human and AI cognition create collaborative intelligence.
Kumar, Swapna & Allen, Megan. "Taking Responsibility for Meaning and Mattering: An Agential Realist Approach to Generative AI and Literacy" (2024) Recent application of Barad's framework specifically to AI and literacy education. Demonstrates practical applications for current challenges.
Latour, Bruno. Reassembling the Social: An Introduction to Actor-Network-Theory (2005) Complementary framework for understanding how agency distributes across human-technical networks. Useful context for Barad's more radical claims.
Pickering, Andrew. The Mangle of Practice: Time, Agency, and Science (1995) Historical analysis of human-machine interactions in scientific practice. Provides concrete examples of how agency emerges relationally.