[fiction] Future fashion

A tale of Generative AI in the fashion industry.

Today is September 20, 2026. After many attempts, I finally get the chance to interview the two entrepreneurs who 3 years earlier create Iconic, a fashion brand recognised for its innovative designs and quality materials. On my way to the hotel in Milan where we are to meet, I could not help but think: this story looks more like a fairy tale. Two sneaker lovers with no prior background in design and experience in the fashion industry decide to launch a brand and 3 years later they are leading a unicorn; those companies that are valued at more than $1 billion. In the tech industry, reaching that valuation is not easy but definitely achievable (less than 1 in a million start up eventually do); what impresses me is they did so in the fashion industry. An industry known to be slow and difficult to enter. But somehow, they did it.

I jumped out of the robotaxi and dashed to the reception. After announcing myself at the desk, I was rushed to their room. Andreas Müller was the one to greet me. He was rather upbeat and hasty in his movements. While waiting for the other interviewee, we started chatting about the Milan Fashion Show, and he explained how it was important for him and his associate to showcase their latest collection and definitely cement their brand as a legitimate fashion brand. “There are still some doubters, you know.” On those words, someone knocked on the door. Amina Kamara, the other member of the duo rushed in. She excused herself for being late and sat next to me on the sofa while Andreas was sitting on a chair across from a table. Both young in their early thirties, they sported casual clothes. Not very fashion-like.

“So Phillip, tell us… what secrets can we share with you?” Amina questioned, in a cheerful tone. Andreas gasped and leaned forward as if he knew the real interview was starting.

“Well, it is an understatement to say that you moved the needle in your industry. As you know, on my blog, I interview innovators like you who do things out of the ordinary. And there is one question I need to ask: what triggered you to start your company?” I said.

Andreas looked at her and explained “Ami and I met during our MBA studies. We realised that we share the same passion for sneakers; but shoes out their look rather the same and the quality is not there anymore. So we decided to try and create a limited selection of shoes that one could wear without encountering a dozen of people with the same pair.”

“I am sure you know it already, neither of us had a background in design let alone any knowledge of the fashion industry. I spent my last 7 years as a Machine learning researcher at MIT. Andreas is quite the genius when it comes to video games design. But nothing related to apparel,” Aminata added.

I noticed that, Andreas gasped slightly when his colleague alluded to his prowess. Maybe a nervous twitch. He leaned backward and crossed his legs.

He then continued, “To be completely transparent, the first person we hired was Melissa, our chief designer. But we did not hire her to make new designs. She’s got on the team to help us understand how all that shoe design thing works. Her last job prior to joining was to source material for a clothing brand. I can tell you she really knows her stuff.”

“She has the title, but we know who is the real boss when it comes to create new items. Our algorithm, which you might have heard of, is the true designer here. We interact with it to create lots of iterations on a product we want on the market. We have spent the first two years building it and training it so that it can create new cool stuff,” she added.

She continued, “If you think about it, human creativity can be euuuh limited. If you ask a designer to make you a brand-new pair of shoes, chances are he will come up with something not far off what he has seen in the past (modulo alterations here and there). The machine, more or less, has the same limitations. But sometimes, it ‘hallucinates.’ It does things that you and I would not expect. It surprises us with things that are beyond the unexpected.”

Andreas nodded, and said, “That is the creative, artistic part. And to be completely frank, the first iteration of designs the algorithm create looked mmhh ‘different’ from something you would want to have on your feet. But we were willing to give it the time to get where we wanted. And for this, we don’t really expect the machine to explain how it came to this suggestion. The second part, when it comes to the structure of the garment, that is where interesting stuff start. You see, there are more than 30 different parts in a shoe. Each can be made with different material which will affect the look and feel of the end product. That is when Melissa and her team intervene. With the preselected designs, we can discuss with our algorithm the details of the look and feel we want to the model mmh the shoe. Sometimes the design language is ‘youthful’, so we need a sporty look to the pair. On other occasions, we decide that a shirt should be as breathable as possible with a fabric that fits this bill, with the style we want to it. At this stage one of the most important parameters we give the program is the total cost we want for the cloth. From there, it is able to make the best possible choices and suggest iterations within that cost range.”

“We don’t worry that much with the material because we are connected with our suppliers and we know — I mean the program knows — what types of material they can provide and what is hot on the market. We have encoded the structural behaviour of the materials in the program, so it pretty much knows how to mix them to create the best pieces,” she added.

“Wow, Interesting!” I said at her in utter amazement. “So that where the whole process begins? What happens next?”

“The whole thing is done in our 3d studio, so we can interact with the product and simulate things. Our models can try and predict the wear of the piece overtime or how a long exposure to the sun can affect the colours. We can pretty accurately tell how a pair of shoes will look like after a year or two; or how it will react to a long exposure to water. This helps make some adjustments. We really want our pieces to be as durable as possible. It is part of our beliefs; they should last as long as possible. But we know there is a trade-off to it. Durability is one of our core mmmh signatures. We also know that people are more forgiving with non-natural material if we tell them why we make this choice: Why we go for faux-leather instead of calf-skin for example.” said Andreas. “From there, we can 3d-print the candidates or ask our manufacturer to prototype 2 or 3 variations to see the product in the flesh. The process usually takes less than a week.”

“Yes, we go very fast at that point because the competition can pick up cues of what we are going for and try to go past us with a similar product.” she said. And then sighed “That’s how this industry works. Manufacturers works with all sorts of brands. So designs, you know, can be copied easily. But we are careful who we want to work with.”

“From there we can compute how many garments people will buy. This is possible though trend and past performance analysis. We are very careful not to inundate the market with pieces and be forced to offer discounts down the line. As you can see we are very much driven by the Data we get,” she said.

“I was not expecting less from engineers,” I added. We all had a good laugh on those words.

“But I believe where we pull the right move is our ‘style advisor’. Let me explain… you can go to one of our stores to get your body scanned in the 3d pod. It will create a 3d version of you: a ‘digital twin’ like we like to call it; along with other metrics it captures such as your preferred colours and materials. From there we are able to advise you on the best sets (of clothes) suited for the day taking into account your calendar, the weather, etc. We are working on a new feature which will allow you to add photos or references of your clothes even if you bought them from another brand. And all this through 3d. You can explore also explore our range from the comfort of your living room either through your computer screen or via virtual reality googles.”, said Aminata. “This strategy allows us to also have a very low rate of return [2% versus 30% in the wider clothing industry]. People know what they are buying because they tried it on virtually. We have also heard that people preferred to resell the items they do not want to keep, often at a higher price on the second hand market instead of returning them. We closely monitor those resale circuits because we don’t want to erode our brand value, you know.”

At that point, I had a aha moment. These people are all about capturing Data to complement their customer knowledge; and lock them (the customers) in. Clever, but scary! They will know a lot of stuff about people. Are people really willing to share this much about themselves?

“People love this feature. Me, I do. I don’t spend that much time trying to find the best outfit for the day. I lead a fashion brand, but I have to confess I am no fashionista,” she added jokingly.

We all laughed again. At least my interview was going well.

“There is another aspect that I would like to touch upon before my time with you is up. I have recently realised that the content I am seeing on my mobile app and the webpages are different form that of my wife. I mean even the product descriptions are not exactly the same. Sometimes she gets pictures when, for the same product, I get more text and more details. How do you do that and why?” I questioned.

Amina jumped to the occasion, after a quick look at her colleague. “That’s because the program knows you are of the ‘analytic’ type while your wife is more of a visual person.”

I got starstruck on those words. How do they know that? Yes, my wife is more engaged with visual content, but… wow!

“How do you…” my question was cut short by the duo, giggling. They were expecting me to ask this question, of course.

“mmmh sorry we interrupted you. The app knows the preferences of you and those of your wife because you told it. Well not verbally but you gave it some cues. From your profile and interactions with it, the app tries to understand how you like your content. In your example, it guessed correctly that you like to know what you are going to buy, through and through. Overtime, it adjusts your preferences to serve you with the best possible information in a good format. The content is generated on the fly: the program decides whether you will be happier seeing a video showcasing the product’s design or an image with cool details about it. It is able to do so because, remember, it was the one who designed the product and it knows what format (videos, text, 3d) suits you best and when. For your wife, it understood that she is not into the details but is more appealed to the aesthetics. We want our customers to have a pleasant time on the app but also in the stores. We work very hard for you to have the same type of customised experience online or in our stores… for you only.” he explained while she was nodding continuously.

On those words, their assistant came in, telling them it was time to head to the fashion show. The pair excused themselves got up and followed the assistant hastily through the door. At least we have had time to shake hands. Andreas stop at the door and turned to me as I was still trying to close my bag and get out too.

“It was nice meeting with you, Phillip. I am a big fan of your blog. Please let us know if you need more information… And and feel free to patch a draft through our assistant, as we agreed, before sharing with you audience.” he said.

I went to the lobby and sat to assemble my notes. It now downs on me how they are able to maintain such a low headcount — 43 people, with all functions and store accounted for. Their extensive use of technology in core functions such as design or sourcing is their secret recipe.

Now that I know the Chief designer is in fact a program, I have a new perspective on the brand. I am, at the same time, nostalgic of the time when we knew there was a face behind a creation and excited about This new type of creativity. And nowadays programs and robots are everywhere. Did not I just ride a robotaxi to get here? Who knows, maybe in ten years, from now programs and robots of all sorts would be running entire companies on their own.

Those two entrepreneurs have harnessed the wonder of zeros and ones to create products that were both unexpected and delightful to their customers. And that’s what people spend their money for nowadays. The people I talked to about the brand we more exited by the products themselves than the faces behind them. People do not even know who Mr. Müller and Mrs Kamara are.

It made more sense now to me that such a small and young company would have reached this unicorn status, so quickly. It was not in fact the products per se that were glowing in eyes of the investors but rather the potential of their software. If their goal was to make a point about the potential of technologies, once again, I believe they have done so. I can imagine them launching a new company, with the same type of software to design all sorts of products: cars, planes, buildings, etc. And this without ever needing experts help. Is that the end of human creativity?