Introduction

Contemporary digital media, often understood as the transmission of information and the exchange of symbols, and the associated theories have undergone dramatic development with the emergence of AI. This blog explores the rapidly growing perspective of technology, including generative AI, and connects it to Ingold’s theory of clay and wood, which are considered materials in his theory. How is it that AI, in which even emotions and ideas are digitally “encoded,” is able to behave like a material that “co-creates” with us? We apply Ingold’s “correspondence” as a theoretical framework, using our own experiences writing and editing with AI as a concrete example. Through these explorations, we hope to reconsider AI as a new “digital material” or medium, rather than treating it as a mere tool or black box, and to encourage its discovery and development.

Theoretical frame work

Ingold’s concept of “craft” is not simply a technical handiwork or task. For him, craft is a process of correspondence between humans and materials, a way of thinking and cognition that “knows the world through making.” While it’s commonly believed that knowledge and form are acquired after a material is completed, Ingold’s theory is different. Rather, it emerges during the interaction that occurs during the process of creation. Therefore, focusing on the material’s properties and resistances, and the process of responding to them are crucial.

Ingold also uses examples of working with materials like clay and wood to demonstrate that each material has its own affordances and constraints, which create resistance. Artists, specifically saying that potters, do not “give” clay a shape, but “discover” it by receiving its response. The response Ingold advocates here refers to the properties of clay, such as deforming when pressed and hardening as it dries. What he is trying to argue is that in this dialogue, the creator’s hands move as if they are “listening” to the power of the material, and that therefore production is an act of collaboration rather than domination.

His theory, which argues that “materials themselves mediate,” further supports this idea. If we consider that materials themselves, such as clay or wood fibers and their texture, mediate the relationship between humans and the world, then mediation does not simply involve the “transportation” of meaning, with the material world simply receiving it. But rather is the act of existence and recognition being collaboratively formed, actively influencing the generation of meaning. This understanding of the “mediacy of materials” makes it possible to fundamentally reexamine the common notion in media theory that media are technological channels for transmitting information.

From Material to Medium: Rethinking AI

At first glance, AI would appear to share nothing at all with Ingold’s clay or wood. It has no smell, no resistance, no texture. But when we use an AI model to write, generate images, or brainstorm, we notice that it behaves less like an inanimate pen and more like a living material.

We can ask ChatGPT to write a paragraph, but it rarely writes exactly what we intended. Its tone either overshoots or undershoots; it introduces unexpected twists, or it stubbornly misinterprets. The process becomes iterative: We adjust the prompt, clarify the request, reject, re-ask, and build from what it offers. The dialogue is not unlike how Ingold describes craft—attending, adjusting, and responding to the material at hand.

AI is not raw material, but far from inert. It “pushes back” in its own affordances and constraints. A large data-trained model has tendencies, biases, and styles that we must work with, just as a sculptor works with the properties of marble. If Ingold were writing today, we think he might view AI as a type of “digital material”—a medium that demands attention, negotiation, and responsiveness.

The Co-Making Process

Working alongside generative AI reveals that making has more to do with coparticipation and less with control. Coparticipation happens at several levels:

1. Iteration and Resistance
When the clay slumps or fractures, so also do AI outputs often fail. The “resistance” is stylistic or semantic rather than physical. Our job is to adapt—amending our input, redirecting, or embracing the surprise.

2. Unpredictability and Surprise
Ingold highlights how makers tend to be surprised by what happens. This surprise is magnified in AI. The algorithm taps into patterns that are invisible to us, and the outcome can be lovely, infuriating, or creepy. But it is here that new knowledge comes into being.

3. Shared Agency
Ingold would oppose isolated human authorship. In creation aided by AI, authorship is yet more openly dispersed. We bring purpose, provocation, and judgment. The AI brings trained statistical relationships and probabilistic imagination. What we get is neither all ours nor all machine-made—it is a joint artifact.

A Personal Example: Writing with AI

When I use AI as a writing or brainstorming tool, the activity is not so much that of typing on a blank page but rather that of joining a studio discussion. Suppose I am attempting to come up with a theme for a fantasy essay. I put a rough concept into ChatGPT. It returns with a partially clichéd but partially stimulating outline. I seize upon a phrase it generated, flip it on its head, and follow it where it goes. Then I give feedback to my rewritten version, and the model suggests edits.

What amazes me is how learning is achieved through this process. I do not receive ready-made answers. I learn things by being in touch. Ingold speaks of making as a kind of thinking in action, and I am living that on the internet. The “thinking” is not internal to my head. It is distributed between me and the AI system. The medium itself is integrated into my mode of thinking.

Broader Implications: Rethinking Media

Ingold’s remarks also lead us to ask what we mean by “media.” If we take his argument seriously that materials themselves mediate, then AI is not just a platform where communication occurs—it is itself a medium.

Clay facilitated the connection between potter and vessel. AI facilitates the connection between digital representation and human imagination. The “material as media” idea is now applied to the algorithmic.

This has ethical and cultural implications. If AI is an agential medium, then producers must pay attention to how it shapes effects—through biases within data sets, through the aesthetics it values, through the forms of image and language that it authorizes. Just as a weaver is sensitive to the nature of threads, a digital producer has to be sensitive to the nature of AI outputs.

Why This Matters

For us, applying Ingold’s model to consider AI changes our entire perspective. It keeps us from thinking about AI as if it were a magic black box or as a neutral computer. Rather, it is more like wood or clay, something which requires skill, patience, and sensitivity to work well with.

At the same time, it extends Ingold’s ideas. Making is not confined to physical substance anymore. It can happen in virtual, algorithmic space, where “material” consists of data and statistical inference. But the rules remain the same. Knowledge emerges in the dialogue between maker and medium, human and material.

Connection to other theories

This perspective also resonates with other theories that we’ve read in class. Semiotics teaches us that signs mediate meaning, but Ingold reminds us that material does. Critical Terms in media studies are often about representation, but Ingold re-centers the process. Evocative Objects suggest that things are given meaning by personal and cultural associations, but Ingold—and we would argue, AI—suggests that things are also given meaning by their becoming, by the process of being made.

AI invention highlights the limits of a purely symbolic or representational conception of media. It shows that mediation is not just about the transmission of messages but co-production with material—material that may be wood, clay, or code.

Written by Mio, Rai, Saber