I think it began in pre-pubescence, as a desperate love of futuristic objects. I lay blame on each clack of the typewriter Gene Roddenberry fortuitously crafted episodes of Star Trek on. Who would not explore alien worlds "to boldly go" on board the Enterprise?
But there is quite enough blame to go around. Might I mention Arthur C. Clarke and Stanley Kubrick who wrote 2001: A Space Odyssey? And while on the subject, lest I not forget, Ridley Scott's Alien.
Moviegoers and TV viewers alike witnessed the trials and tribulations of deep space voyaging, but their popularity resulted in a consequent dearth of forward-thinking visuals in favor of trite characters and plot serialization. Maybe I’ll cut a bit of slack to George Lucas' Star Wars franchise and a nod to his conceptualizations of Death Stars and anti-gravity vehicles.
Fortunately, the famine of creative futuristic imagery is starting to subside with nourishment from AI generative artists. And there is more in common with music and such art than you might imagine at first blush. So while I laugh with unconstrained hysterics at the villains possessing human physiology adorned with alien heads, I have to admit it has taken generations to reimagine a madly repulsive Xenomorph as birthed by H. R. Giger in the 1979 masterpiece Alien.
I had to learn more about AI artistry. So I scoured cyberspace to understand where futurism might "boldly go" and found the artist known as Synthography Art. Prepare yourself for exploratory teleportation via wormhole into foreign worlds born of the human imagination.
On Process and Creation
Joe Caplan: When you're generating AI art, do you ever think of it like composing music – layering elements, adjusting tempo, finding harmony?
Synthography Art: Absolutely. Creating AI art is similar to creating anything – music, writing, cooking; no different from any other art form. Your vision comes to life through a process, thinking about composition, thinking about elements, thinking about the overall feel, experimentation.I try not to plan too much as invariably your vision will change, perhaps a little, perhaps a lot, perhaps to the point your initial vision is nowhere to be found in the end and you find yourself somewhere you never imagined you would be.
JC: Does your creative process have a soundtrack, or does music interfere with the kind of focus AI art requires?
SA: Absolutely. The soundtrack may be literally playing in the background or playing in my head that can steer the outcome of the image, sometimes not music at all but more like a simple beat or sound effect that can set the tone or mood for the prompts, and [the] final result.
JC: Have you ever generated visual art directly from music – feeding lyrics, album covers, or sound waves into the AI creator? What happened?
SA: To be honest, no; with so many of my own ideas piling up in my head, that is not something I have gotten around to yet. Having used some AI tools to create “music,” I have in turn had album covers generated in the process, and some of them turned out well in representing the lyrics. I decide to do one today; here it is:
On Collaboration and Control
JC: Is working with AI more like conducting an orchestra (you guide but don't play every note) or like improvising jazz with a partner you can't fully predict?
SA: When generating AI art I find there are three main roads.
One road involves a specific vision that I am trying to realize which very much is like composing music, layering elements and finding harmony to achieve precisely what I am looking for.
Another road is the exact opposite – I am purposefully being vague as to elicit something unpredictable.The last road is a combination of the two, where perhaps I am prompting for something either predictable or unpredictable, and then add something unpredictable.
JC: When the AI surprises you with an unexpected result, does it feel like a musician hitting a wrong note that turns out to be brilliant?
SA: Absolutely! Often the prompt and associated generative settings are engineered to do just that!
JC: Do you think AI art and AI music have the same creative challenges, or are they fundamentally different?
SA: The creative process in creating AI art vs AI music for me is very similar, but each form of expression has its own different challenges and nuances.Some differences I would point out from my experience are:
One’s personal knowledge of visual art vs music is a large piece of the puzzle. Having gone to art college and studied color theory, drawing, painting, art history [and so on] and having indulged in photography, 3D animation etc. has been extremely helpful when prompting images, that is, [techniques like using] complimentary colors, media types like watercolor or oil, composition, 3D lighting types like volumetric lighting or spotlighting, cameras, lenses or film speeds for grain.
I own an electric guitar, a ukelele, a handheld Casio keyboard and a kazoo, and I am a master of none. By comparison, my knowledge of music leads me to feel like I don’t really know what I am doing when prompting music, so I find it much more difficult to get the sound I am looking for, versus [creating] an image.
AI art is visual, and AI music is audio, obviously, but when it comes to curating the music or imagery there is a big difference. One can curate hundreds if not thousands of images in a day with AI art but with AI music you must listen to the output, which limits the number and quality of your curated pieces you can produce.
On Emotion and Authenticity
JC: Can AI-generated art evoke emotion the same way music does, or is something lost without human intention behind every choice?
SA: I believe so. In some cases images are created without intention or a combination of chaos and intent.
I asked ChatGPT about automatism and its definition is: “In fine art, automatism is a technique where the artist begins without conscious intention – making random marks, scribbles, or paint splashes. The idea is to let the subconscious mind guide the initial marks. After something interesting appears in the chaos, the artist may then switch to intentional work to develop the image.”
Often times it is impossible to tell if something has been created with AI or not, often times it is impossible to tell how the image was created at all, [whether] watercolor, photography, print etc.
In the end you're just looking at an image. Any image can evoke emotion.
JC: Have you created AI art that feels like it has a "mood" or "key" – something that would pair perfectly with a specific song or genre?
SA: Indeed, post-apocalyptic or cyberpunk pieces pair very nicely with punk rock music. Some of the Weekend Retreat [series of] pieces I have made make me almost hear classical music in my head when viewed.
JC: Do you think people will eventually have favorite AI artists the way they have favorite musicians, or is that relationship different?
SA: I don’t see why not. When seeing AI art in my feed, I sometimes know straight away who made it. You become familiar with different AI artist’s styles and subject matter. Some you may like more than others, just like any other art form.
On Innovation and Tradition
JC: Music has centuries of theory, notation, and tradition. Visual art does too. Does AI blow all that up, or does it need those foundations to be interesting?
SA: I don’t think it blows anything up. Did photography blow up oil painting? Did the synthesizer blow up music? Does it need those foundations? Does it need to understand music theory? Perhaps not, [but] it does need to listen to music to train.
AI music is interesting but I would imagine [it would be] more interesting if it has a clear understanding of the theory, notation and tradition, if that makes any sense.
AI music will never replace humans performing live music. AI music creates AI music. AI music is a genre unto itself. AI music/art is just another tool in my mind, not unlike a pencil or kazoo.
JC: Are there "genres" of AI art emerging the way music has genres? Do you identify with one?
SA: Yes, of course. You can see genres of imagery that never really existed before as a result, like “Cursed AI” or as I mentioned before, forms of automatism. Personally I am a huge fan of liminal AI images and publish to a liminal group regularly.
JC: If AI art is a new instrument, what's the equivalent of learning to play it – what's the skill that separates a master from a beginner?
SA: Not unlike a musician having several guitars or keyboards, having different models at your disposal is useful when you are looking for a certain “look.” Some [AI] models are amazing, some not so much, but the real trick is to learn the tools that do the generation using these models. You can use powerful online tools like Midjourney or generate images locally on your computer. There are powerful open source apps like ForgeGUI or ComfyUI offering complete control over image generation. With the myriad number of settings, the exact same prompt can yield very different results. Plugins associated with these tools are endless, so the learning curve basically never ends.
To answer your question, the skill that separates the master from the beginner is: Knowing what tools are available to you.Exploring and trying new ones constantly.
Knowing what the settings in these tools are and precisely what they do.
On Cultural Impact
JC: Do you think AI art will have "movements" like music had punk, grunge, or hip-hop – even moments that define a generation?
SA: All art has "movements," so I do not think AI art will be any different in that regard. AI art is accessible to most anyone, so I see a large and diverse demographic of AI artists making for a large and diverse number of movements that certainly could define not just a generation, but also define a generation in high detail.
JC: Has a piece of music ever directly inspired an AI artwork you created? What was the connection?
SA: I cannot think of a particular music piece per se; however, when exploring new prompts I often have music playing in the background, and this has a noticeable effect on my prompting and image results. If I am listening to M83 versus the Sex Pistols or Ella Fitzgerald, it sets the mood. Different words tend to find their way into my prompts, directly affecting the output.
I will say I love all genres of music, as long as it’s good.
JC: If you could score a gallery show of your AI art with any musician or composer (human or AI), who or what would it be?
SA: At this particular moment I would likely choose…OK, I just spent far too long thinking about this, as a lot of my AI art [involves] cyberpunk, robots, post-apocalyptic or ancient civilizations and ruins…let’s go with Nine Inch Nails; that could change by tomorrow though.
On The Future
JC: Five years from now, do you think people will care whether art or music was made by humans or AI, or will the quality be the only thing that matters?
SA: I certainly think some people will care whether it was made by AI or a human, but probably less than when AI art first came out. In the end, AI art is simply another form of artistic expression.“
"Beauty is in the eye of the beholder” – Margaret Wolfe Hungerford, Molly Bawn, 1878.
It may be impossible to tell if an image is AI.
JC: Could there be an AI art piece so moving that it needs to be experienced in complete silence, or does AI art demand a soundtrack?
SA: The Mona Lisa, a Rembrandt, the statue of David, the Taj Mahal? I do not think any art needs to be experienced in complete silence, but for some, one may prefer not to have any distractions when taking it all in.
AI art can be moving. It does not demand a soundtrack in that it can stand alone just like any other form of art, I would think.
That said, music certainly can work symbiotically with many forms of art and does so in a very special way that no other form of expression can.
JC: If AI could generate a perfect synthesis of your visual art and a musical composition in real-time, would you want that, or does the separation matter?
SA: It’s called photo to music [or image to music] and is freely available online.
I just tried one out for fun just and this is the result:
Stone and time, carved in stillness
Ancient voices echo in the hall
Water glimmers, deep and endless
Whispering secrets to us all
Shadows dance in the colonnades
Candles flicker, gentle flame
We gather on these drifting stairs
Each searching for a name
[Chorus]
Sanctum of light, cradle of hope
Ripples reflect the longing inside
In circles we move, drawn to the glow
Sanctum of light, where memories reside...
[Verse 2]
Granite stories underfoot
Worn by lifetimes, silent prayers
Turquoise waters hold the truth
Soft reminders of years and cares
Arches open to the sky
Statues watching, solemn, wise
We find ourselves within the ring
Lost beneath these ancient eyes
[Chorus]
Sanctum of light, cradle of hope
Ripples reflect the longing inside
In circles we move, drawn to the glow
Sanctum of light, where memories reside
[Bridge]
Here we remember what time can’t erase
A moment suspended, a sacred place
Beneath the surface, so vibrant, so clear
The waters invite us to draw closer here
[Chorus]
Sanctum of light, cradle of hopeRipples reflect the longing inside
In circles we move, drawn to the glow
Sanctum of light, where memories reside
[Outro]
Candles melting, shadows grown
Echoes fading, steps unknown
But in this circle, we remain
Drawn to the sanctum, again and again
On Philosophy and Identity
JC: Is AI art more "honest" than human art because the AI has no ego, or less honest because it has no lived experience?
SA: This is a tricky question. AI art is derived from training on human art and images and in a way represents humanity as a whole. An AI artist will input a prompt and tweak settings and perhaps do some post-processing to images. I do not think you can separate human input from AI output. AI art can be just as honest or dishonest as human art, because it is essentially also human art, and one can certainly inject their lived experience into a prompt.
JC: When you create AI art, do you feel like an artist, a curator, a conductor, or something else entirely?
SA: All of the above.
Is prompting an art? I think so, but I am one of those people who believes everything is art. Each word and its position in a prompt has an effect. Prompts can include many other instructions, from defining word weights to cycling through different words for each prompt. Prompt results will all change based on the model, app settings and plugins you might be using. The fact is, prompting can become very complex and take hours, days or weeks to perfect, if at all. Curation is very important. Deciding which images to keep will essentially define your style.
Conducting can literally be visualized when you look at tools like ComfyUI where you “wire up” complex systems to produce what you're looking for.
JC: What's the AI art equivalent of a musician's "signature sound" – how do you make your work recognizably yours?
SA: The most important thing is to make art for yourself, not for anyone else, not for what you think people want to see, but what you want to see. Follow your instincts while continually creating images and your style will undoubtedly come to the surface over time, it will change over time, you cannot predict what you will be doing in the future, but you will be pleasantly surprised when you find out!
Wild Cards
JC: If your AI art could make sound, what would it sound like?
SA: Underneath the ruins, dystopia, post-apocalyptic, wasteland, pits to hell, secret bases under volcanoes, I can hear a low level hum…from some far off source of power.
JC: Have you ever wished you could "hear" the training data the way you can see it – would AI art sound chaotic, harmonic, or like nothing we've heard before?
SA: All of the above, I would imagine.
JC: If music and visual art are both just patterns and math, does AI prove they're the same thing, or does it prove they never were?
SA: Everything is art. Everything is math. Everything is relative.
To see more from the artist, please visit:
www.synthographyart.com
https://www.instagram.com/synthographyart
https://www.facebook.com/synthographyart
Images courtesy of Synthography Art.
0 comments