Pre-verbal Optical Pleasure
Survey
November 2025
The reason I chose this image - made back in late-2022 with Midjourney – is because it connects me to a very primitive aesthetic response, an almost pre-verbal optical pleasure. When I was six or seven, I had a handful of collectible Thundercats cards issued by Panini in the UK and while I’ve never been an acolyte or collector of anything, I still have a visceral memory of the materiality of those cards: foiled, iridescent, holographic. I’m not sure if any visual artefact has ever produced the same response in me, but this image from 2022 somehow revives it. Yet because it was so easy to produce something so precious, there’s also a sense of loss involved in this image: the bankruptcy of optical pleasure and the end of a longstanding materialistic relationship to the visual. It’s also a really dreamy pony.
With text-to-image workflows particularly, there’s a sense that we’re mediating through an opaque interface, one that requires alchemical incantations and a much more intuitive way of steering latent space.
How do you see the relationship between artists and their generative tools?
We’re used to sliders, settings, drop-down menus – the empirical software interface that privileges a certain model of expertise. With text-to-image workflows particularly, there’s a sense that we’re mediating through an opaque interface, one that requires alchemical incantations and a much more intuitive way of steering latent space. This can be highly personal and that make creative practice a fecund space for new metaphors – fishing, driving, sailing, surfing... maybe even horseback riding! The perspectives that interest me most are to do with dimensional transpositions – the quirks of mapping n-dimensionality within a human brain unused to that modality. That’s where things get cosmic.
How do you assess standard generative images aesthetically? What formal features stand out?
Ok, here’s another metaphor: generative AI was a big stone dropped in a pond, and now there are many ripples and hundreds of ducks flying everywhere - lots of moving targets in terms of formal aesthetics. At a time like this there’s a need for a deeper understanding of image technics, and I suppose this is where I fall back to thinking about CGI. Today’s GenAI machinations have a lot of precedence in the way that CGI developed between the years of 2000 and 2017. The ‘AI winter’ corresponded with a ‘CGI summer’ when techniques for rendering simulated phenomena improved significantly. That’s something being repeated now with Gen AI – the same metrics of photorealistic representation, the same modular approach, but according to a different technical rationale, one that’s more mysterious and unstable, but equally aligned with a larger sociotechnical project. What I’m saying is that viewed from a distance, the ducks migrate in fairly predictable patterns.