A regularly updated statement on the use of AI.
I’ll start by stating that this wasn’t written by AI. It’s just me making sure that I clearly explain my stance on generative ai when creating imagery (Midjourney, Stable Diffusion, DALL-E, Leonardo, etc. etc.).
How I Use AI:
Let’s start here. As a veteran designer and artist of over 25 years, the advent of AI creating images using a text prompt is facscinating and chilling at the same time. When I started seeing Midjourney (the G.O.A.T. for the time being) having really, really amazing results around version 5, I started to get really excited. What started as a pressure test for the system- writing a ton of unhinged prompts to see how creative/stupid/offensive/ridiculous/amazing ended up becoming a part of my daily workflow for inspiration and ideation.
When I’m in design mode, I am constantly in research mode. Searching the web for color, tone, angle, lighting, inspirational references… you name it. AI image generation is a fast track to getting some great references for the work that I do. That work is done by my hands, and my tools. Those could be anything from the Adobe suite, to pen and paper, paint, collage, you name it.
AI will never replace the profound experience of taking pen to paper, hand to mouse, finger to fretboard – and making something real.
When I post AI generated images, they’re being posted to show a part of my process, or to discuss an overall tone, style, or topic. At Phantom Workshops, we’ve expressly specified that there will not be any AI-generated content for sale, ever. Fuck that.
AI image generators are great for helping a process, or speeding up a workflow. They’re marginally good at solving problems that may have taken way more time and effort. They also make really fucky mistakes, and they’re only guessing, even when your prompt is a paragraph long.
How I Don’t Use AI:
It’s a bit of a concern that there are people that are using AI with all of the stops pulled out to create imagery that they then sell commercially, especially when they are using prompts that directly include the specific names of artists. Not cool, and not really my thing. Artists that have had their work scraped without consent should have the right to choose to have their content ignored/removed from these models.
Work created by AI should not be able to be copyrighted (this is a super-confusing part of the conversation and is a huge pain in the ass to figure out depending on your stance). Artistic styles (AI or otherwise) cannot be copyrighted.
We’re Confident That There Will Be A Renaissance
Without a doubt. People will feel the fatigue from this new form of image generation, and will want to get to a point where they can hold the original in their hands. The smell of the paint on the paper, the texture, the story behind the work, the heart, sweat and tears of a true human skill immortalized on physical media.
Don’t steal people’s work and claim that it’s yours. Stay human.
There’s way more to this. I’ll update it when appropriate.