Lysxis’s interview — Revel in AI

The Lysxis Interview

A conversation with AI Artist Lysxis, aka @Lysxis on X

INTERVIEW CONDUCTED ON APRIL 24, 2025

ALL IMAGES BY LYSXIS ARE GENERATED USING MIDJOURNEY (NIJI) UNLESS STATED OTHERWISE

Introducing Lysxis – The Dreamer in the Shadows

I’ve always admired Lysxis, a British artist whose work walks the knife-edge between dreams and nightmares, memories and myths. His visual world is populated by delicate characters—lost boys, haunted girls, nameless misfits, and, at times, androgynous figures—who seem to drift through ethereal landscapes, pose for us and the camera, or dissolve into the very shadows that surround them. His images often feel like pages torn from a forgotten fable or memories half-remembered from a fever dream. There’s a quiet intensity and shyness to his characters, many of whom stare right back at us, wearing expressions of melancholy or muted wonder, as if they’ve just glimpsed something vast and unknowable. At other times, the trauma becomes real, and lost children seem to be victims of Pied Piper-like figures who torment them and might unleash a world of pain upon them.

The recurring motifs are important in Lysxis’s work—red eyes glowing in the dusk, antlers crowning anonymous silhouettes, strange spirals etched into flesh or drifting across backgrounds. These elements add layers of symbolism to his images, whispering of pain, identity, and transformation. But even in the stillness of his compositions, there’s tension and unresolved suffering: a soft hand curled around a smartphone or coffee cup, a lone figure submerged in water, children standing before—or perhaps running from—an all-seeing cosmic eye. These are portraits of shy defiance, where fragility and darkness share the same skin, and innocence is ripped from the young.

Lysxis’s technique is as unusual as his themes. He works almost exclusively with Midjourney’s Niji mode, a model designed for anime and Eastern-style aesthetics, but he twists its bright conventions into something altogether more uncanny. The colors are muted, and washed-out tones elevate the sense of disquiet. His prompts are often lines of poetry or surreal fragments of writing—dreamlike cues that yield raw, unpredictable visuals. This poetic approach shows that his work is less about perfection and more about feeling—about glimpses of a tender, broken world trying to make sense of itself.

In Lysxis's universe, symbols matter—eyes, dice, antlers, halos, spirals—and so does ritual. Some of his characters appear mid-transformation or mid-ritual, caught in some alchemical moment between boy and beast, light and void, or simply trapped by nefarious forces whose intentions are unclear—be they ghosts from the past or shadowy figures from the present, poised to become future tormentors. But Lysxis always brings it back home, where, in the comfort of his many “Good morning” posts, his subject simply poses and sips black coffee while the world ends outside. Ultimately, Lysxis's work is a study in subtle rebellion: a rebellion against silence, against conformity, and, most importantly, against forgetting. I am honored to feature him in this interview.

Let’s discover Lysxis together.

Good morning everyone

Without getting too personal, can you tell us a bit about yourself?

I have so many preconceptions about other artists on X that it seems a shame to disabuse people about me! I've been mistaken for all sorts of things, which is usually fun, but I will say that I'm English, male, and an archaeologist.

Could you please tell us which country you live in?

I live in the UK.

What led you to begin working with AI imagery?

I was doing some research for a story I was writing, and kept coming across pictures with the same artist's watermark on them. I liked them, so looked the artist up. It turned out they worked with AI, with a thing called Nijijourney. I wondered what that was, and if I could do it as well. My early efforts were embarrassing, as I didn't understand how to prompt at all. But one thing led to another, and here I am.

What AI tools do you use?

Niji almost exclusively. I've tried to get to grips with Midjourney since I started, but I'm not really interested in the trend toward photorealism so I always end up finding ways of making the results look like what I do in Niji. I did this again with v7, and it dawned on me - if Niji is the tool that gives me the results I want, why not just use it rather than try to replicate it using something else?

I've been experimenting with animation using Higgsfield and Kling lately, but I don't feel like I have enough control over them to find the results satisfying.

What inspires you?

All sorts of unmentionable things! Plus a great many of the excellent artists I follow, mythology, history, literature, music, dreams, the world out there, you name it. Feedback from other artists and the friends I've made in the community keeps me posting.

LIGHT & DARK

(One of @revelinai’s favorite images by Lysxis)

In which other medium, if any, do you practice art?

I used to work as an illustrator and a technical artist, but I haven't done either in many years. I'm a writer and a musician, although music has become only a hobby as my professional life has got busier.

Would you consider AI-generated art true art?

It can be, but it isn't by default. In the same way that not all verse is poetry, not all writing is literature, not all sound is music, some AI generations are just pictures.

It depends, I think, on what you mean by Art. With a capital A, Art is an aspiration, an intention, and those things reside in the artist - the tool is merely the thing the artist uses to bring the aspiration into the world - what does it matter what the tool is? What the aspiration is harder to pin down, but for me at least it's the desire to express the inexpressible, to illustrate metaphorically some part of the ineffable. That sounds pompous, and I'm not about to claim that my Boys in Bathrooms do that (although to me, privately, they do). That, in fact, illustrates something else about art - that art that only speaks to the artist doesn't achieve the status of Art - there has to be some degree of communication, of a shared experience with an audience, however small. Traditional Arts can be exclusive, and elitist, and set about by shrieking gatekeepers. Anything that democratises access to the ability to create and experience is okay by me.

In these terms, arguments that AI-generated images can't aspire to Art are, frankly, stupid. We've understood this since at least the 1930s and the various reactions to The Work of Art in the Age of Mechanical Reproduction, let alone to synthesised music, sampling, Gysin and Burroughs' cut ups, and any number of other examples where the craft lies somewhere other than where traditionalists insist it has to be. Issues surrounding intellectual property, copyright, theft, and the monetisation of the creative process by profiteering corporations are another matter, but those aren't new, either.

Please share one or a few of your favorite images with us.

Here are some of my favorite images:

Untitled

Do you title your AI-generated art? If you do, what inspires you to come up with these titles?

Sometimes. It depends. If I have a very clear idea of what I'm trying to produce, and pursue that successfully to a conclusion, then I may well title the result. But I am in love with chance, serendipity, surprise, and mistakes, and the results of those things I tend to leave untitled. Sometimes I won't title a piece until I post it, and often those titles have less to do with what I was feeling when I made it than they do with how I'm feeling when I post it. Sometimes I omit the title from pieces that have one - I often don't like to limit the possibilities of how people might interpret my pictures.

When do you tend to be the most productive, and do you work in long sessions or short bursts?

I'm at my best in terms of initial image generation first thing in the morning. Anyone who's suffered my 'Good morning everyone' pictures will probably understand the importance of lots of strong black coffee in my life. I work on refining things throughout the day - I'll often come back to a thing numerous times over several days. Having said that, other pieces are just prompt, count the fingers, done.

Untitled

What type of prompts do you prefer: text or AI-generated descriptions through fed images?

I'm 100% text, initially at least. I tried images early on, but I never liked the descriptions the AI generated, or the results if I used them as prompts.

When prompting with text, do you write simple text-based prompts or complex ones?

Both, and everywhere in between. I have prompts that are a single word, an sref and a p code. I have prompts that are hundreds of words long with multiple nested parameters. I've yet to satisfy myself completely about how much of a prompt the AI reads - I'm still experimenting with that.

Do you think text-based prompts should be shared within the AI art community?

I think that's a matter of personal choice. It can be fun trying out other people's prompts, certainly, and can help to push you to places you might not otherwise go. On the other hand, when I first started out, someone sent me a prompt out of the blue. I guess they'd seen me struggling and taken pity on me, and at the time, I was very grateful for their kindness. In retrospect, I rather wish they hadn't done it, because prompting - and working out what makes a good prompt, and why, and what works and doesn't for one's own goals - is the most of the skill of our craft, and it took me a long time to work out my own way of doing things. Which is a long-winded way of saying 'maybe'.

Then again, in Niji at least, the prompt only does so much of the work, and the heavy lifting is often done by the sref or p. Someone once said to me that using someone else's p was like wearing their underwear, and I think I agree.

What is the most unconventional method you have used to create an image?

I'm not sure what's unconventional because I don't know what other people do. I'm sure I'm fairly straightforward. I often generate roughs in Niji 4 or v7 and then rerun the results in Niji 6, which sometimes results in a nice effect. Maybe the way I prompt? I sometimes see other people's prompts, which are very different from mine - mine sometimes tend to be very abstract. I'm a writer myself, and I follow some writers and poets, and some of them are kind enough to let me use their words as prompts. I love using pieces of writing as prompts that don't make sense as prompts - the images that come from those are the basis of what I consider to be some of my best pieces. I don't really use other people's srefs - I have several of my own, and if I have anything resembling a style, I think that's why.

How many images have you generated using AI technology?

I don't have the faintest idea, but it must be tens of thousands.

astragalomancy

Please share your X (Twitter) handle with us.

@Lysxis

Tell us a bit about your X handle and X username.

When I first started, I intended to use AI art to tell an illustrated serial story. My ambition was several orders of magnitude ahead of my ability, so I failed at it very quickly. My username was an amalgam of parts of the names of three of the characters in that story. As people who've followed me for a while will know, I've been shedding vowels like a tree loses leaves in Autumn.

How many X followers do you currently have?

Currently fluctuating around 2400.

How many X accounts do you currently follow?

820. I always feel bad not following back automatically, but my timeline is already too big for me to give everyone the attention they deserve.

How long have you been on X (Twitter)?

Since June 2024.

Icarus 3

Is there any other online platform where we can view all of your previous work apart from X?

No. I'm on Bluesky and Rodeo, but only X has everything.

In this world of cryptocurrencies and non-fungible tokens (NFTs), are you currently selling your digital art on any platform? If so, which platform are you using?

Well, I have some things for sale on Objkt.com. I have no real expectation of making money from my art, but I like to support other people by buying their work and it's nice if I can fund that at least partially by selling my own.

Are you selling tangible AI-generated artwork, such as high-quality prints like Giclée prints?

No.

Could you suggest some AI artists you like that we could follow?

I always assume that everyone I follow has 65K followers, and it's a constant surprise to find that many of them actually have about the same number as me, or fewer. Here are the first few of those on my timeline this morning.

Every one of them is worth your time, and I'd also like to add @tismjj, if they ever get their account back.

Is there anything else you would like to add or share with the AI community?

I'd like to thank everyone for being so welcoming and accepting. I very often feel like a fraud, so it's been great finding such an open community. Let's hope it can continue.