This month marked the transition from administrative preparation to active research, with the completion of 2 patient and 2 clinical interviews. This represents the visible tip of a massive iceberg of preparation. Because of the sensitive nature of Thyroid Eye Disease (TED) and its potential impact on facial identity, collecting clinical data requires an ethical framework that goes well beyond standard compliance. To ensure full participant agency, I developed bespoke information sheets and formal consent protocols detailing how these personal narratives will inform the work. 

To safeguard sensitive qualitative data, I designed and built a 100% offline data ingestion pipeline on my local machine. The technical architecture utilises WhisperX for high-accuracy local transcription, Pyannote for speaker diarisation, and Microsoft Presidio to scrub personally identifiable information (PII). I also developed a custom, local side-by-side interface that allows me to manually verify speaker as before the output is saved as a structured JSON file. This human-in-the-loop finessing is something that in this case is important for what will happen to the words next.

" Is this part of the art?" you may ask - Yes, and also it's part of the care for the conversations. The local database preserves precise timestamps, enabling time-based data manipulation in the digital layers of the final installation. But more fundamentally, this painstaking offline infrastructure represents a commitment to digital stewardship. I feel that I am the custodian of these conversations. Each one has been so precious, and every interviewee has been incredibly honest and generous with their thoughts, feelings and imagination. All the effort to keep everything anonymised offline is not just fulfilling a legal obligation, but is an active ritual of compassion.

I'm having my preconceptions and prior assumptions challenged by these interviews. And I like that. I've always maintained that the most important part of people-based qualitative research is to be constantly able to be surprised.

To maintain clinical and ethical rigour throughout these dialogues, the interview materials were formally reviewed and approved by Mizgin Arslan, a BACP-accredited psychologist. Mizgin serves as a vital grounding presence and a safeguarding consultant throughout this project. We have established a schedule of regular supervision sessions to monitor the psychological safety of both the participants and me as the researcher-artist. Mizgin's background in somatic therapies and active creative practice makes her an invaluable clinical anchor as we navigate these sensitive personal histories.

On the technical front, Dr Dylan Banarse, Senior Research Engineer at Google DeepMind, has agreed to serve as my technical mentor on this project. Dylan has been working on AI since way before the rest of the world caught up to the possibilities. Our bi-weekly, 45-minute check-ins focus on translating qualitative data into generative parameters and tackling fundamental questions like "can computers really make art?" Dylan is patient enough to answer my most basic questions (including what I imagine are the same ones he gets asked at every dinner party) and an active contributor to creative coding ideas - prompting me to push the technology in new directions and try out new things. His advisory role is supported by Google DeepMind, and he provides invaluable technical oversight and creative ideas as we explore how to translate linguistic patterns into tangible art.

Long before I experienced any personal health issues with my thyroid, I had a deep fascination with clinical illustration and medical imaging, particularly the historical, hand-drawn diagrams that predated photography. Part of my proposal is to research the archival collections at the Wellcome and UCL. Preparing for this has forced me to articulate what I am searching for, and why. The old anatomical and instructional drawings stem from a time when medicine was not so specialised and not so divided from art (think of the single origin of the word “theatre”). Perhaps I feel that at that level, it makes it more accessible and more visceral. I’m also curious about how other cultures see the eye medically, and physiognomy - that distasteful strand of the history of medicine saw faces as deterministic.

Next week, I will commence archival research at the Wellcome Collection, having secured my research library credentials, alongside upcoming work at the UCL Joint Library of Ophthalmology to trace these early visual representations.

In the physical studio, material trials are moving at a slow, rhythmic pace. I am currently developing a tapestry weaving sampler. Tapestry is an intensely low-technology medium. While my initial strategy was to let data analysis directly dictate weaving instructions, my discussions on AI have shifted this focus. Although the most abstract layer of digital work remains binary, the sheer volume of processing required to execute a simple LLM prompt is mind-boggling. We are looking beyond simple binaries to the physical density of the hand-weave to represent the weight of human experience. I have also initiated a rug hooking project using the project's cool, clinical colour palette and am working on drawings that translate visual metaphors described by patient participants.

Finally, I visited two major textile exhibitions in London: Textile Art Redefined at Saatchi Gallery and Threads at Sarah Myerscough Gallery. The works at Sarah Myerscough stood as powerful, physical presences through their scale and as a collection that utilised materials ranging from tyre rubber to natural and foraged fibres. The curation felt coherent and understated, whereas the Textile Art Redefined featured many different kinds of textile works and whilst all were individually excellent, I felt it was an attempt to provide something for everyone, which diluted the impact of each. This reinforces my ambition for The Gaze Interrupted: to bypass the need to constantly reassert the medium's seriousness and instead focus entirely on scale, presence, and conceptual weight. 

One interesting work from the Threads exhibition was a collection of perhaps 40 different woven ‘baskets’ in forms resembling parts of the body, all thoughtfully arranged and well lit on a single panel. What struck me about these was that they were small individually, but presented together, they were as impactful as the larger single objects. The works at Threads were sculptural and relief-based rather than painterly, and I think this could be something for me to think about in my work moving forward.

I am pleased to report that my first formal mentoring session with artist Heather Phillipson took place early this month. Heather’s mentorship is the backbone of this project, helping me to ensure the creative integrity and the emotional impact of the outputs. This is in no way a footnote! Her agreement to mentor me and presence on the project are absolutely key. Choosing the right moment to announce has been holding me back. Our first meeting was a long, therapeutic walk through the wilds of North East London, where I explained this project and some of the work I'm planning for this R&D year. Our next meeting has been pencilled in for the last week in June, and I will write more about this key strand then.