Looking at AI art through the eyes of a Senior designer
Marten used to draw sneakers on paper, now he uses Midjourney
Hey Marten, we discovered your work on LinkedIn and we thought it was inspiring! Your LinkedIn headline is quite impressive! You are a Senior designer and art director working for one of the biggest Dutch design agency and you are a jury in both the Webby Awards and the Lovie Awards, You even won a design award recently, congratulations for these achievements! I think we will have a really interesting conversation about Design, digital art and AI. Can you tell us who you are and how you ended up doing AI-generated art?
Marten: I’m Marten, a designer and art director from Amsterdam. Since almost 5 years I’m working at DEPT®, a digital agency with more that 4000+ people across the globe. How I ended up doing Generative AI art? Well, a friend of me said; “You definitely should check out this new thing called Midjourney, you will love it.” And he was right. Within no time I was paying for the most expensive membership and downloaded the Discord app on my phone. I never got addicted to something that fast. It felt like not even the sky was the limit…
It seems you are kind of obsessed with fashion, most of your AI content is around sneakers, clothes or fashion related. Can you tell us why you think these are great use cases for GenAI?
Obsessed is a big word but I definitely love playing around with fashion items in AI. I think it’s because I’m a designer. And as a designer I’m interested in a lot of creative topics like architecture, art and also fashion. Especially Nike sneakers is something which I love to experiment with. At a very young age I already was sketching sneakers on paper and later also on the computer. I liked it but I was limited because of my skills back then.
Also I’m a big sneaker lover, I’ve tens of pairs and I always would have liked to design real sneakers myself. Now it’s that easy and fast to create whatever you like, I went a little viral with some sneakers I generated. In the end I created hundreds of sneakers and got in contact with custom sneaker designers who actually are willing to create my AI generated sneakers in real life. A childhood dream.
You’ve been designing products for many years now, I am curious to hear your thoughts around GenAI in terms of work. How does it fit ( or not ) in your workflow, what are the current limitations and best ways to leverage Ai?
In my opinion Generative AI is really ideal to use within your process for moodboards and sketches. When I’m curious how something will look I test it within minutes. Also it’s ideal to brief photographers or illustrators with by showing them some examples created with Generative AI. If I had to do it all by myself it would take way more time and that would be a waste. But there are definitely limitations. The quality is not high enough, hands and eyes aren’t always going well, the details are a bit wack sometimes and a lot of times it feels like I’m not in control enough about the output. Generative AI always surprises me which is a good thing, but when you’re looking for something specific you don’t want to get surprised too much…
What is your POV around the big crisis currently happening around pro-GenAI and against-Generative AI? Especially Thinking about the art station polemic for example.
Well… A lot already has been said about this in the past year. Call me naive but I’m pretty optimistic still. Let’s use Generative AI within our process to help us with the little tasks but let’s respect the craft and artists. We need them.
Are you more into Midjourney, DALL-E 2 or Stable Diffusion? And what about chatGPT?
I’m a designer so I definitely would go for the tooling where I can generate images. I absolutely love the ‘inpainting’ and ‘outpainting’ features from DALL-E but I already fell in love with Midjourney from the beginning. It delivers the most artistic images in my opinion and that’s what I love to create the most. It doesn’t have to be photorealistic for me. What I also like in Midjourney is the Blend mode where you can blend up to 5 images which results in crazy visuals. What I do like about DALL-E is that it’s not integrated into Discord, which is sometimes pretty annoying with Midjourney.
What do you think is the next big thing that’s gonna happen in the next couple of weeks/months, ai-tech related?
That’s a good question. I guess that will be generated videos and 3D. I already saw many initiatives and companies working on these and I think this will get bigger and bigger. Don’t forget that tools like Midjourney didn’t exist a year ago, and look where we are now. The future is bright, but the level of brightness is in our hands.
Subscribe to our newsletter!
With your experience using the tools, you probably discovered a couple of tips and tricks, Which ones would you be ok to share with our audience?
I’m always using short prompts, in that way the output surprises me the most and via ‘upscale’ and ‘make variations’ I end up with some great results. Takes a bit more time but I hate it to get limited or restricted within my imagination and exploration phase. And another tip is using words like ‘surrealism’. No matter which prompt you have, this word adds some magic to it. For sneakers words like ‘Octane render’ worked pretty well.
What is the next thing you are going to try using Ai tools?
I’m still playing around with the Blend mode. Via this feature you easily can create collabs between brands but also could create weird patterns. When you’re blending 5 images the original images you used as input are not recognisable anymore in the output. That’s also what I really like to play with. Besides this it’s nice to add my own designs (created in design tools) to this and create fun stuff with it.
Is there anything else you want to share with our audience?
Keep on playing but respect the artists.