This is the second article by Fiona Norton on the intersection of Artificial Intelligence and Running. Fiona reminds this writer that human interaction is needed for artificial intelligence to be successful in these endeavors. AI, as Ms. Norton notes, is a tool.
AI and Running, Part 2: AI Affects User Research & Product Design, by Fiona Norton
Artificial intelligence is changing the way that we design and create products. While it’s still in the early stages of development, programs like GPT-4 show a promising future for new ways to get insights into large amounts of data. AI is beginning to and will continue to affect user research and product design in the shoe industry.
User research is the less talked about but a very important part of UX design. With any product, a group of people researches what people want in a product and what they like or dislike in the prototypes leading up to that product. One UX Designer, Jasper Kense, predicts that AI will automate UX Research. “A UX researcher can then focus on asking the questions and not doing the research. A researcher can give questions to the AI, who then asks a large sample size of these questions. The variety and quantity of data would be far superior to what we design with nowadays.” This means that the AI would get and organize data from people, which is one of the most time-consuming parts of the job. This would allow UX researchers to spend more time actually analyzing the data and determining how it can be used.
Shoes, created by Fiona Norton, using DALL-E, an AI software.
Image recognition software can go through non-text documents like pdfs or jpeg images and derive text from them. This is going to drastically change the way that companies sort through and organize documents. Natural Language Processing, or NLP, can help by sorting through and learning from the information we already have, via text and now even images, on what people have bought in the past, what they liked about it, and what they wished was different.
It can provide sales insights by determining who’s buying the shoes and what their needs are and identify discrepancies in how shoes are currently being marketed vs. who’s buying them. For example, most running shoes are targeted towards runners, but if more everyday workers (nurses, cashiers, baristas, etc.) are buying them, then there’s a discrepancy in what the shoes are being made for vs. how they’re actually being used. A nurse buying running shoes to stand on their feet all day may not be worried so much about how their gait is affected but more about overall comfort and relief. By separating out running shoes from everyday work shoes, each group can receive more direct attention to their needs in the development of the product, and more time and energy can be put into developing running shoes specifically for runners.
A shoe company could use AI to organize their database so that every shoe they’ve ever made has been scanned and has a 3-D model with information about the materials used, how expensive it was to make vs. sell for, color scheme and design details, etc. In addition to organizing databases, AI can help in the world of idea generation and product mockups. Image generation software like Dall-E or Midjourney can understand specific written requests and create an image that represents that. I’ve used Dall-E as a fun way to create some interesting hypothetical design collaborations between shoe companies and high-end fashion designers, musical artists, and style trends for Gen-Z. Because Dall-E is an open software not specifically trained for designing shoes, the results are not extremely polished. However, they do show a great understanding of what different shoes look like and how elements from all kinds of inspirations can be transferred into the design process.
The way I see this technology really benefiting companies is to save time and money by using these AI prototypes to test what kinds of shoe designs people are drawn to before actually making them. I was inspired by Sydney Stark’s work with AI shoes and color palettes and used that as a jumping-off point to create an example of how different shoe brands could use AI to prototype new shoe designs. The examples below showcase how New Balance, Nike, and Adidas running shoes might be designed for different generations using color palette trends that represent their individual tastes. This technology has the possibility to level out the playing field of competition among running shoe brands, so long as the smaller brands start using it now and finding their niches in product design and marketing with AI.
3D Printing combined with 3D image recognition could make bespoke running shoes a very real possibility. Dr. Scholls has had the technology for about a decade to create a “foot map” and determine which custom insoles are best for you. These are just inserts, and they’re not fully customized to your foot specifically, but they’re a good indication of how similar technologies have already been implemented. In an ideal bespoke product, a company could take a 3D image of your feet, a heatmap of how you disperse pressure, and a video of your gait. AI could either take an existing product and tweak it to fit your specific needs or create a new product entirely and create shoes that are custom formed for you. A technology like this could benefit groups needing more specialized attention, such as people with mobility or orthopedic issues.
The overall benefit of using AI in user research and product design is that we can create more efficient shoe designs for different consumers. By automating the user research process, we can better identify target audiences and what their needs are. Image generation and processing in the world of product design can help brands better cater to the different needs of people.
For more examples of AI-generated sneakers, check out the following links:
AI Shoes Pinterest – A Pinterest board of AI shoes I’ve created using Dall-E by OpenAI
Sydney Stark – Shoe AI – A collection of Nike shoes created by using NVDIA’s AI technology
This Shoe Does Not Exist – AI shoes generated using Nvidia’s AI technology
Please read our other columns on Artificial Intelligence and Running: