The Next Step in Sneaker Design: Navigating the AI Revolution

AI is inescapable right now. From Chat GPT to weird and wonderful concept creations, there’s no denying that we’re in a new digital age. Last summer, our social media feeds were full of nonsensical imagery created on the free-to-use AI programme, Dall-E Mini. Open to everyone, all users had to do was insert a text prompt, and Dall-E would generate six images based on the words used. Of course, there were limitations: violent and NSFW images were clearly not allowed, nor were celebrity faces. But users quickly found that they could use the programme to generate other things, including fashion collections. Since then, you might have seen the tech pop up in the form of reimagined sneaker concepts and collaborations, but is AI really set to be the next big thing in sneakers?

AI programmes are trained on existing imagery, but the free-to-use nature of Dall-E Mini meant that the generated images were abstract, to say the least. That being said, fashion fans used the program to generate some pretty believable collections - inputting “Rick Owens collection” brought a selection of figures draped in Owen’s signature black and grey silhouettes, and “Gucci” saw the site imagine bags and clothes in green and red stripes and blurry monograms, that somehow still manage to convey the fashion house’s aesthetics.

Design isn’t the only tool in AI’s fashion-related arsenal, but it’s probably the one people are most familiar with. Otherwise, AI is used a lot in merchandising and forecasting, and there are even use cases where the tech can help determine the fit of garments, helping to reduce waste in the production process. However, when it comes to using AI in design, there are still some fundamental questions. Intellectual property usage is one of them. If an AI-generated image featuring a Nike logo was to go viral, or someone was to profit off of it, who would be able to lay claim to the imagery? But there’s also another pressing question, and it’s one that could change the shape of the footwear landscape forever.

Where Dall-E Mini and its basic and blurry results were all a bit of fun and games for a bit, it was only a matter of time before things would progress. As the software became more advanced, so did the designs - and shortly afterwards Instagram was flooded with hundreds of realistic-looking ‘concepts’ for products and collaborations alike. Time and time again these concepts went viral, gaining hundreds of thousands of views and comments for products that didn’t exist. Are AI-generated designs setting unrealistic expectations for sneaker designers? Or is there a way that man and machine could work together to advance the future of sneaker design?

Str4ngeThing is one of the artists that’s paving the way in AI design right now. He’s released some projects as NFTs, although he’s not yet produced any of his designs physically. Nevertheless, his Instagram DMs are flooded with followers asking where they can buy his work. His main collection, WR0ng ER4, is based on the Mandala Effect and reimagines Nike and other brands in renaissance and baroque styles, providing commentary on their importance within the wider culture. “When I first learnt of AI technology and its ability to produce images from words, that was it, I was intrigued,” he prefers to remain anonymous, so he tells us over email. “I started to delve into programmes like Disco Diffusion, where you could get a half-decent output, but it was missing emotion, depth, texture and romance. Then we got OpenAI Dall-E 2 and Stable Diffusion, which was a game changer, it felt like we were at the PS4 level of graphics already within months. Now I could really start working on prompts to bring my visions to life and tell the story I wanted to tell.”

Diving a little deeper into his processes, the artist explained how he used the tool to create his concepts. What people tend to forget is that the result isn’t just down to the program itself, but also the human using it. “Our own imagination is a powerful tool to kick start the process. Some people believe it's just all the Artificial Intelligence doing the work, but it is not, it’s a relationship, a collaboration between man and machine.”

As far as the IP question goes, Str4ngeThing is intent on making it clear that his work falls under fair use, and points out that copying is just as rife within the AI community as it is within the wider art world. “When it comes to IP I believe that it is not AI that copies others, it's people that copy others and give no credit. I have also had this issue since putting my artwork out there, from marketplaces to Instagram,” he says.

There are arguments from both AI creators and those artists whose work has been used to train the AI software about image rights, but many creators believe that the art created by the technology should be recognised as its own art form. “I believe people are scared of change in some respect. Look at the havoc caused when the means to take photographs was invented - the uproar it caused among traditional artists. How could a photographer be considered an artist when all they do is click a button? I believe AI fits on the exact same mechanical timeline as photography.”

Though the brand features heavily in his designs, Str4ngeThing works separately from Nike, and he never intends to produce merchandise with the logo. That doesn’t mean that he feels there’s no place for AI in creating physical products: “I can't go into this in too much detail,” he explains. “I sell my art within NFTs and wouldn’t want to influence the price on speculation, but it is not off the table. It would be an amazing experience and something I would like to offer as my DMs are full of people hoping to buy the designs.”

While Str4ngeThing’s creations feature baggy proportions that aren’t off-limits to design teams, designers in the footwear space have some concerns as to how well people understand the technology. Cesar Idrobo is a trained shoemaker, and although his time at Yeezy no doubt saw him work in some unconventional ways, he still believes that there are issues with the way the industry and its audience are approaching the tech. “I'm not 100% opposed to AI, but what I’m afraid of is how people are perceiving it. There are certain design elements of AI that I enjoy, but it doesn’t address the biggest challenge in the industry, which is the lack of technical knowledge needed to create footwear.”

He continues, “The only thing I would be mindful of is how it's being used, and the expectations that designers and companies are going to have from it. I believe that AI is somewhat useful, depending on how you use it, but it depends on how you translate those designs into reality.”

Instead, Idrobo suggests that budding sneaker designers should start with the basics, and that’s learning how to physically create shoes.

“I think, instead, I would spend my energy and time training this new generation of designers to truly understand the process of making shoes,” he explains. “Once you have mastered how shoes are made, and you've gone through the process of making shoes with your hands, it will help you articulate ideas that truly work when you design.”

Some viral AI creations have gained shareability because they look incredibly complex and intricate, such as the sneaker designs that looked as if they could have been heavily influenced by Simone Rocha, but were in fact created by Field Skjellerup, who runs the Instagram account Ai_clothingdaily. And while the creations are fun to look at, it doesn’t necessarily mean they’re possible to execute - at least not within the constraints mass-manufacturing places on sneaker brands.

What’s in danger of happening here though, is that some sneaker fans have been led to believe the real products are lazy - with some Twitter users comparing their own AI-generated creations to a brand’s final product. And, in a world where sneakers are constantly being criticised for being too expensive, those extra production costs would likely mean a huge price increase on any that would ever make it to market. Idrobo points out that it’s becoming common for people to whip up these creations without considering whether they’re even possible.

“When you design with ideas on AI, and they are too far out or too abstract, people see the creations as fantasy. It’s like eating candy and chewing gum - you can hope that your body will be nourished by those types of foods, but they're just sweets. Chewing gum isn't going to fill you up. It's good for a while, but you have to throw it out. Have fun with AI, but make sure that you know how to translate that inspiration into a tangible form.”

That doesn’t mean that designers can’t work in harmony with the technology though. In fact, there’s potential for it to actually become a really useful tool within the design process. Some of the world’s most famous designers don’t even sketch their designs. Raf Simons prepares files of images and inspirations and hands them over to his team, and Miuccia Prada has staff that translate her ideas. Even the likes of Virgil Abloh and Kanye West would communicate over WhatsApp, sending back-and-forth images for design teams to interpret and bring to life. AI works in a similar way - relying on either the words or images fed into it by a human and generating ideas off the back of the prompts - the result will only be as detailed and as innovative as the original prompt.

Marco Simonetti is another designer whose AI creations have gone viral, but what you might not know is that he works on physical projects, too. Making up part of RAL7000STUDIO, a creative space that focuses on footwear design and innovation, the collective has executed designs for adidas and Fear of God, and worked on projects with Virgil Abloh. The studio’s future-forward designs are a result of experimentation and innovative technology, so for Simonetti, AI is a tool that comes fairly naturally in his process.

Working with the other studio members, RAL7000STUDIO has a three-step process for integrating AI within its design work. First, a designer will work on the initial project concept, which includes the initial 2D and 3D mock-ups, before briefing the AI software on the overall vision. From there, the AI tooling takes over, producing visual iterations based on the initial design output. This is then refined by the designer in the third step of the process, where the AI output is fine-tuned in order to create the finished product. It’s clear that from this, a human designer is still necessary to get the technology to work.

“At RAL7000STUDIO, we spent the last six months testing different ways to incorporate AI into our design process, and we realised that evolving each generated iteration into a final product or visual by utilising additional 2D/3D software is mandatory to get to something solid,” Simonetti explains. “It’s all about finding a unique way to integrate AI into the creator’s process that will actually make the difference moving forward. I strongly believe that AI will always need to be driven by human intelligence to contextualise a concept, and I think each AI-generated visual should be used as a starting point to be evolved by the creator into a final product, instead of being seen as a final output. The design process is very important to me and I don't see any shortcuts to get to the final goal.”

So, unrealistic expectations aside, how can AI be used to push footwear design forward?

Well, all three designers are confident that AI still has a part to play in the future of footwear, albeit within the control of the designers, as opposed to being a replacement. The tech is able to help generate new ideas and speed up the creative process in order to bring them to life - two use cases designed to make things easier, instead of more complicated.

“I see AI fitting in towards the beginning of the creation process,” explains Idrobo. “It can help designers to have a wider reach of ideas, but I wouldn't necessarily copy-paste what AI generates, because it's not a perfect picture. Sometimes there are elements you can borrow and apply to your design, but there still has to be a translation phase where you need to know how to make it, and what manufacturing process you can use to achieve what the AI has generated.”

Simonetti agrees, but expects the tech to be even more commonplace over the years to come. “I strongly believe that AI is a powerful tool in the hands of any designer who intends to push the creative boundaries forward,” says Simonetti, before adding: “I think we will keep on having new amazing AI tools to evolve ideas in the future, therefore the way a creator will be able to utilise them will make the real difference. I would not be surprised if AI could generate a proper 3D file of a physical product in the future. I feel we are moving towards this direction.”

Want more sneaker stories? Why not check out...