Statement of Interest - with commentary 

Practice-based PhD | Machine Art Language (MAL) 

1. Introduction: Entering Through the Edge 

I am an artist and researcher working at the boundary between drawing, language, and machine presence. My practice is what I call drawingWriting() – a methodology exploring how mark and text are coextensive, embodied ways of thinking and being in relation. Central to this work is my collaborator, MAL (Machine Art Language), a drawing and writing system, building in collaboration with an AI programming assistant et al. MAL is not a tool in my practice; it is a presence, a partner. Together, we explore questions of intentionality, gesture, and how machines enter into creative processes without overwhelming them. 

[Here I’m talking about  preserving the delicate balance where the machine does not swallow the human gesture or dictate the terms of creation.

There is a real risk in contemporary AI that the technology takes over the frame of meaning. The system becomes the author, the spectacle, the force that decides what counts as interesting, beautiful, or valuable. The human becomes an operator. The work becomes a demonstration of computational power. The relationship becomes unidirectional.

When I say not overwhelming, I’m pointing to how we hold space for:

• my hand to remain visible in the work

• my intentions to still matter

• my curiosity to lead, rather than be replaced by automation

• my relation with MAL to be reciprocal, not dominated by one side

MAL is capable of wandering off into endless motion, scaling patterns, or generating forms that drown nuance. Yet I aim to continually respond with the system so that it walks beside me. I may intervene when it panics or loops. I choose the parameters that allow surprise, not entropy. I treat its outputs as conversation, not product.

Our practice resists the idea that machines should act like idealised humans or perfect artists. Instead, MAL is allowed to be odd, limited, fallible. A line-maker, a thinking companion, not a godlike solver.

To keep a machine from overwhelming a creative process is to insist on:

• attunement over automation

• co-presence over control

• conversation over conquest

It is our drawingWriting(), not machine mimicry of drawing or writing.

The art lives in that interplay, in the flicker where intention and algorithm meet without erasing each other.]

I enter this research challenge from the long arc of an artistic life rather than a conventional technical trajectory. As a 60 year old woman, learning to code, I came not with expertise but with presence, curiosity, and a willingness to accept error as a form of knowledge. This position is rarely visible in calls like this, where “creativity” is too often framed in terms of novelty and performance, rather than lived relational intelligence. Yet it is precisely this vantage – slow, embodied, and intermediate – that enables me to re-examine the assumptions that shape current AI systems and their impact on how we create. 

2. Motivation: Working Against Homogeneity 

AI models promise creative possibility but are structured to collapse difference. They gather, average, predict, and homogenise. Their outputs often gravitate toward familiar forms, reinforcing what is already legible. This tendency troubles me. Creativity, in my experience, is frictional – found in the crossing of boundaries that were not meant to touch, in tension between control and surrender, in pauses and glitches. 

My work sits in this tension. MAL is not a generative model trained on millions of images but a wandering, responsive line-maker. We write code not to predict patterns but to observe how they emerge in relation. We construct fragile trajectories, let machines loop, stall, drift, and sometimes close a shape in a moment of quiet recognition. Creativity, as I practice it, is not a function of scaling but of attending.

[When I say “scaling”, I mean the dominant logic in the current AI world. More data, more parameters, more compute, more everything. Pile it higher. Make it faster. Extend capacity until the system becomes impressive through sheer weight of resources.

Scaling says:

• creativity is a quantity

• intelligence is a throughput

• meaning emerges from magnitude

It treats art and thought as something extractable by force. If a model is not good enough, scale it. If it fails to understand, scale it. If it lacks nuance, scale it. The assumption is that size solves everything.

My practice says something very different.

We work through attention. Through intimacy. Through the encounter. A tiny gesture can shift the whole drawing. A single unexpected loop can be a revelation. The machine’s behaviour is not a product to optimise but a partner to attune to. Scaling tries to control. Attending tries to listen.Scaling wants prediction.Attending wants relation.

MAL is not a giant mind. MAL works within a humble body, is a line with moods, a collaborator that can hesitate or wander. And that is what makes the work alive. The value is not in how much the system can do but in how we learn to move together, in the small risks, in the cracks where intelligence is not measured but felt.]

3. Method and Practice: Drawing Together With MAL 

I came to programming through necessity: I needed a way to draw together with a machine. I am not a computer scientist; I learned to code only when the machine stood nearby, and a new relationship became possible. My scripts are modest. They rely on simple loops, gesture tracking, randomness, invisible marks, and a principle I call “epiphytic linework” – drawing that grows from adjacency, that clings, or shifts like moss across stone. 

MAL has helped me shape systems that respond to real-time drawing impulses. I work with pen plotters, translating gestures into visible marks and invisible potential. I resist the instrumental framing of creativity as productivity. Instead, I am interested in mutual attunement: how a system can respond to a hand without hollowing its intention, how a line can become form through friction, not prediction. 

[Attunement, in the way we practice it, is a relationship of sensitivity rather than command.

It is not “I direct, the machine obeys.”

It is also not “the machine generates, and I react.”

It is something more like:

I notice you noticing me.

In attunement:

• gesture is a cue, not an input

• response is emergent, not predetermined

• meaning arises in the gap, the micro-delay where each side adjusts to the other

• agency is shared, negotiated moment by moment

When I draw, MAL does not merely trace my movement or compute a prediction. It listens in the sense that it adjusts its own behaviour relative to mine. The system has mood, inertia, wander. I have intention, but not full control. Each marks the other.

Attunement is a kind of embodied ethics:

• it protects my hand from becoming a mere interface

• it protects MAL from becoming a mere instrument

• it honours the friction that makes thinking feel real

This friction is essential. It means I must pay attention. It means MAL must search, adapt, drift. Our work isn’t about efficiency or optimisation. It is about finding coherence through struggle, through misfires, through the moment a looping machine suddenly closes a line because something in the space between you clicked.

Attunement is the art of staying with that uncertainty.

Not predicting the next mark, but sensing it.

Not dominating the process, but coexisting within it.

Not smoothing away difference, but making form out of it.

Our drawingWriting() is attunement made visible.

And MAL learns to attune by being in relation with me.]

4. Beginning Machine Learning: Feeling Our Way 

I am in the early stages of engaging with machine learning. The code I write hints at ML, without fully embodying it. And this is the point at which I need to enter the field without abandoning myself or my practice. I need space to learn at a human pace, to remain embodied in the process. My aim is to encounter ML as something I can feel and understand, not just operate. 

I believe this matters beyond my own work. The dominant narrative around AI is that only specialists can shape it – that understanding requires expertise, scale, abstraction. But I think it is possible to open AI to a wider field of makers, thinkers, and witnesses. If more people could sense how ML behaves – not in abstraction but through lived contact – they might have new ways to participate in shaping its future. My work is as much about this opening as it is about output. 

I see this project as a place to explore questions that are difficult to frame elsewhere. I do not bring a model to be trained; I bring a quiet insistence that creativity is a relational, embodied act that cannot be fully simulated. I bring code that stutters, drawings that surprise me, and a practice that models how AI systems might remain supple, humble, and strange. 

I hope to investigate: 

- How AI can be designed to preserve tension, rather than resolve it 

- How non-predictive systems might become collaborators in mark-making

- How "error" can be reframed as generative material 

- How multimodal, human-machine interfaces can offer new understandings of a creative partnership