Let's talk about AI art for covers
There's something that's really bugging me when I read how people debate about AI art for covers. People tend to mix so many things up and then they don't even see their own logical fallacy.
I don't want to talk about the ethics of how the AI was trained here, and just to mention that at this point, there is a more ethical usage of AI by simply using AI locally on your computer like Stable Diffusion (means no AI-theft-company gets paid, you do it on your computer) or using something like Adobe Firefly which was only trained on licensed work where artists got and get paid.
What I want to talk about instead is...
The big discrepancy between creative fields.
Graphic designers, illustrators, people who draw and paint always had and still have the opportunity to just use our music for free in order to promote their next piece of art on social media. Because in order to survive, we have to upload our music to all these platforms.
When was the last time an artist approached you and said: "Hey! I've almost finished my next drawing and before I post that, I'd like to have some custom music that I can share with the post, tailored to my drawing. I'd pay you $500 for a commercial license, how does that sound?"
Yeah, this never happened and will never happen, because they don't do it. So where does the idea arise from, that musicians have to hire an artist to create their album cover for them, if artists don't have to hire musicians to create a piece of music for them?
How can there be a one-sided relationship where one side only has to receive and the other only has to give? It makes no sense.
The next subject that bugs me is that
People are mixing up the different usages of AI
There's a huge difference between me creating an instagram account with AI art, flooding comissioning websites and other platforms with my AI art and thus actively making the life harder for actual artists vs me just creating an artwork for my own music.
Vice versa, it's not the same if someone claims to be a musician and floods Spotify and iTunes with music they generated with Suno and takes money from our pool and makes it harder for us to be found, added to playlists and so on, or if an artist just creates a little piece of music with AI for their next post.
I think it should be clearly differentiated between these use cases because one is harmful for an entire industry and the other is simply not. These two are not the same.
And then there's the accusation of
Low Effort
I hear this all the time: "If you used AI art I couldn't be bothered to listen to your music, because if you didn't even put effort into your cover, I assume you also didn't put much effort into your music".
Of course, this is just the internet echo-chamber, you hear people parroting this exact same sentence and of course all those people didn't coincidentally come up with the exact same phrasing.
But let's think about this for a second too: How many bands with very successful albums did just drop a minimalistic album cover with a black or grey background and some title on it? Quite a few, and nobody ever said "oh wow low effort cover I'm not gonna listen".
The entire debate has become so toxic that people come up with the most nonsensical arguments.
They recommed you just take your camera and take some photos of shaky lights outside that you then further "post process" with PS (aka smashing filters on it), putting your title on it and then being proud of it because you did it "with your own hands". Well, congratulations: You got the same shitty boring overused cover that millions of other people have created too already.
If this isn't low effort, I don't know what is. Or just making some geometrical shapes in Illustrator with a Youtube tutorial because of the lack of visual skills, or some 2005-ish wallpaper texture in the background with some font over it.
You see, "effort" can't really be the metric with which we evaluate album art. Instead, it should be
Intentionality
Let's assume I'm mainly composing epic cinematic sci-fi music. My music is about far away galaxies, colonies in space, big space ships, plasma weapons..
What would perhaps be more intentional for my album artwork: If I spend days prompting the scenery, trying 200, 300 different generations, extract single elements I like, combine them with other things I generate, create a new composition from it, post process it, add a custom title, so that I end up with a huge sci-fi colony on a burning planet in space, surrounded by many stars and space ship fleets, or..
.. if I walk outside into my garden, shoot a photo of the moon and use that instead?
Clearly the first approach is much more intentional. Yes, if you write piano music and your musical piece is sad you can get away with taking a photograph of the rainy street outside of your house. But people make different music in different genres and some of these genres demand a much more powerful visual identity that you won't get with a low budget or low skills.
So I think we should begin to differentiate. I'm fed up with people using AI for professional purposes and contesting with actual creatives in their respective fields too. I'm also not a fan of musicians using AI for their actual music or artists using AI for their actual art. I don't want to guess how much % of something I'm enjoying was made by the artist and how much was done by AI.
But I see none of these problems if someone who's doing 100% of his craft in his creative field, uses AI to do some other supportive work that is not his main craft.
Now I'm mainly looking at this from a perspective of musicians / bands with not too much budget. Of course, if we're looking at a band that's famous world wide, making tons of money with a big label in the back, they'll probably have the funds to hire some pretty great designers that are also fun to work with. And then these designers with a few hundred-thousand followers post about it as well and they record a little documentary and completely turn this into a huge PR boost.
But most musicians won't have this experience. They will hire some random person through Fiverr who couldn't care less about the music, who want to get done quickly in order to get paid for the comission but in the same time debate about their artistic decisions trying to tell you how your artwork is supposed to look like and in the worst case end up ghosting you.
Luckily I didn't have to make such an experience yet, but I've read about it countless times. The one time I tried to hire an artist to make art for me, only 2 out of 5 replied. One of them wanted to have $300 for the artwork for a track that I'll probably make <$3 with, and the other (which I ended up hiring) didn't deliver a single brush stroke in the first 3 out of 4 weeks, paired with ghosting and a lot of excuses until I cancelled the request and got refunded.
The bottom line is, in the low budget section it's very difficult to find professionals who are good at managing themselves and also understanding that they do art for someone else, not for themselves. And everyone who's more professional asks for amounts of money that most musicians can't afford because it will cost more than the music will even pay in a time where people don't buy music anymore and a stream gets you something like $0,00003, if you manage to accumulate XXXXX streams on that single track per year. Artists aren't hiring us either to compose music for their posts, and we're not claiming to be artists, opening up professional accounts and flooding platforms to contest with them and make their life harder.
So honestly, what's the deal? I think it's time to differentiate. Excited to hear your thoughts, I'm really open for different perspectives. But I hope to hear some more logical input than all the logical fallacies that I see repeating endlessly in these debates.