©RossHelen via Canva.com
Can Generative AI Help Retailers Create and Design Better Products?
An online jewelry manufacturer, J’evar, has created its own generative AI application in order to design new products.
According to Digital Commerce 360, J’evar’s jewelry designers utilize a tool that permits the input of details about the product’s composition and specifications, after which the generative AI depicts an image of the product. This innovation drastically reduces the manual design process, a benefit highlighted by Amish Shah, the founder and CEO of J’evar, which specializes in jewelry adorned with lab-grown diamonds.
Introduced in 2022, J’evar’s AI-powered jewelry image generator accesses a comprehensive archive of images, text, and metrics for materials. This database encompasses specific details, such as the weight of precious metals like gold and silver, which the generator references before creating the final image.
For instance, Shah explained that when he wants to create a bangle, he simply provides the generative AI with a description, detailing the gold weight, dimensions, and desired style. He even has the option to request 50 different designs based on a single instruction. The AI then generates a variety of designs, with some being ready to produce immediately. However, certain designs need adjustments from J’evar’s team before they are viable for production. Shah pointed out that the main obstacle to realizing some of these designs is the current limitations in diamond-cutting to match AI-generated shapes. Nevertheless, Shah believes advancements in machinery will soon overcome this challenge.
According to Fast Company, “Generative AI takes AI to the next level. As the name implies, generative AI generates data, whether it’s audio, computer code, 3D objects, video, etc. It uses existing data to create novel output and, when used for product design, generative AI creates new ideas and designs rather than recognizing and categorizing existing data.” ChatGPT is currently the best-known example.
Generative AI is becoming a pivotal tool in product development, particularly during the conceptualization phase. This technology excels at integrating various techniques, processes, and established solutions into novel ideas and structures. Consequently, designers get to delve into previously unexplored design possibilities. Moreover, the results generated by AI can expedite and enhance the design testing phase.
However, Brendan Witcher, vice president and principal analyst at research firm Forrester, believes generative AI still has its limitations. He noted that “it doesn’t think about engineering and structural elements or physical viability yet,” though he also acknowledged that it could eventually get to that point.
Per Digital Commerce 360, “Even with its current limitations, Witcher says generative AI’s value comes from the assessments it already has learned to make.” He added that “it can also come up with ideas humans couldn’t or wouldn’t think of because the human mind doesn’t process information the way artificial intelligence tools can.”
While AI’s functioning mirrors human job performance, it also moves beyond, assessing the best subsequent steps from more data than humans can process. Witcher emphasized that this capability to create unprecedented images or concepts has immense potential.
When it comes to J’evar, Shah believes that generative AI functions as a supportive aid to human designers. He explained that the technology doesn’t dictate when a design piece is complete; that call is made by a human. Drawing parallels with graphics software tools like Adobe and Corel, Shah positions generative AI as another tool in a designer’s kit, not a replacement for human creativity and decision-making.
One such tool is Grid Dynamics, which has already been used by big names like Google and Apple and offers a starter kit that “provides Generative AI models for rapid prototyping of new product designs based on sketches, references, and textual descriptions.”
Discussion Questions
Do you believe there are aspects of design that should remain solely in the hands of human creators, or is there potential for AI to take a more central role? Could generative AI dull the human aspect of product design? When used, how might it impact the overall product development timeline and quality?
Poll
Loading …BrainTrust
Gary Sankary
Retail Industry Strategy, Esri
Nikki Baird
VP of Strategy, Aptos
This is an excellent example of how generative AI can have a significant impact on design ideation. Human designers provide the context, objectives and then selects the final design while the rough designs are created using generative AI. This process would not only significantly speed-up the design process, but also reveal new design concepts that human designers may not ever identify. And while one might argue that this will dull the design skills of human designers, I disagree. It’s still a human designer setting the requirements and selecting the final designs. In this case, generative AI aids in the process.
Generative AI may spark some creative ideas that designers may not have thought of previously. A combined AI and human creativity may be the best approach. It is hard to believe that we will move to solely AI-driven designs without human review and edit. This could help speed up the design timeline and enhance creativity.
Another example of this comes from UK grocery retailer Waitrose. The chain used AI to explore and identify cuisines, flavors, and taste preferences that were trending. This helped them determine that they needed a Japanese range of ready meals. AI was very helpful in sifting through data to pull out the trends. Of course, when it came to confirming and refining the findings, humans had to step in. Humans were also responsible for creating the menus and products – AI does not have taste buds so would not be much help here. A good example of how AI plays a role alongside humans rather than replacing them entirely.
Much of design involves interpretations of existing designs from competitors, higher end products, etc. In that case, AI would be helpful in bringing all of that information to the table more efficiently. For true innovation though, human creativity is still critical to that success.
Generative AI (GenAI) is already way beyond making Van Gogh portraits of your cat playing the guitar. I liken the usage of GenAI to how a CAD tool aids an architect in creating designs, only multiplied a thousand times. As the sophistication of this tool grows, it will revolutionize the field. It’s also excellent at iterations, so we’re not looking at one-shot guesses at product design here, either.
I think the most important thing to remember when using GenAI for product development is, well, humans. Retailers need to make sure that customers are included in the design process. It’s their wants and needs that will translate into purchases later. Also, those involved in the process will naturally become champions for the products it produces. Today, GenAI hasn’t learned the limitations of the materials and tools to produce its concepts, but very soon it will. That’s when we will see a creative explosion unprecedented in our lifetime. This will get interesting fast.
This is exciting stuff. Generative AI should be able to have a huge impact on both the speed and breadth of ideation. In apparel, the sheer volume of new styles required season after season means that the design process carries a heavy load. The demand for differentiation and distinction is constant, and sometimes overwhelming. So if generative AI can both speed up the process and produce more options to edit from, that’s a big deal. Notice that I said more options to edit from, not more choices at the end of the process. Retailers don’t need to be offering more choices. They need to be offering more differentiated and distinctive choices. If generative AI can help reach that new level of differentiation, great. Product development is all about Data + Design, and generative AI can now have a role in both of those elements. But ultimately, the best storytelling will be a result of human curation of the final assortment.
GenAI seems to work best – for now – when it inspires human creativity. In a world of cognitive load from switching contexts, I think where GenAI really shines is in helping someone shift gears faster. What I’m watching, though, is where usage leads to a potential sea of sameness. If everyone is using it, not as inspiration, but as THE design, how will you come up with something truly unique?
While I agree Generative AI has many uses, I am scratching my head at this one.
If you provide generative AI with a description, material information, dimensions, and desired style and then choose from 50 different options, you cannot claim you designed that piece of jewelry. The AI did. And it will recommend those designs to others so whatever you make can’t ever be exclusive to you.
Personally, I want my designs (articles, copy, etc.) to be things I truly created, not things created by AI that I merely tweaked.
That is why we go to trunk shows and meet with designers. To create something unique and one of a kind, not copied from what has been done before. So true!
For creatives, inspiration comes from everywhere. Using AI to help jumpstart the creative process to get to that “next great idea” sounds like a terrific use case for this technology. The idea is that a creative can filter for a specific idea or form they’re interested in quickly, which is a game changer, in my opinion.
I once visited a vertically integrated fast fashion Retailer that had a team of 4 designers who were expected to create a minimum of 15 – 20 new models per day each, for the buyers to select from. They were cutting/pasting and building samples at a rate that really didn’t look like much creativity was being infused into the potential new products. Can this be done by AI today? surely it can. Net new innovation with forward looking charactoristics…..maybe not today…coming soon – I’d bet on it.
This is a classic conundrum. AI excels at drawing from vast pools of data, both obvious and obscure, to design products that serve people’s current unmet needs. But humans are far better at taking those illogical, hunch driven leaps of creativity and intuition to design products that no one knew they needed. I think AI would have given Henry Ford the genetic code to design faster horses, not the idea to invent a machine to replace horses. Going forward the obvious answer is that we need both. Humans will be needed to work in concert with AI, and still tasked to find that idea to spark a product we didn’t realize we couldn’t live without. Only now they will have an AI partner to do the iterating and refining into a scalable product that meets environmental, safety or other requirements and regulations.
I think most agree that generative AI is an incredible tool with unlimited potential that typically adds value in a number of ways, including design. It can provide a number of inputs to the design process, but the final output should reside with a human designer. Nothing is as creative as the human mind, especially, when considering human-based characteristics such as situational awareness, empathy, touch, feel, context. This is where human art of design becomes far superior to a tool.
In regard to the jewelry sector, creativity is in the eye of the designer. To create a piece that is collaborative with the customer. Inspired from having a dialogue at a trunk show or taking a piece of jewelry that has an art deco look and making it modern for example but still maintain its heritage. For fast fashion, I get it. For jewelry in particular I would suggest that designers through their own training and experience generate art that is unique, distinctive and inspirational – let alone one of a kind that the wearer embraces as representing their sense of fashion and artistic inspiration.
It will have a larger role for sure, however not necessarily because it is better, but because it is cheaper…A sad trend in my view.
Imagine you’re sitting across from a high-priced consultant so that you can ask questions and get opinions on anything related to retail, including product design. You may or may not like what your consultant says, but you feel comfortable knowing you had a conversation and received feedback and ideas. That is what generative AI can do for you. However, it costs nothing close to what a consultant would charge. If you can ask the right questions, you’ll get the answers you need or ideas to consider. This technology has been developing and improving over the years, and this is the year it’s made it to prime time. It is available to everyone for little cost (and even free, depending on the resource). The future is now!
Love it! Appreciate that humans are still the ones approving/checking the final designs. There’s a scalability here that’s fantastic and that I believe consumers would be as excited about as the brand.
Ah, the race to mediocrity accelerates. Why do we expect so much from AI? It seems to feed directly into management desire not to have to rely on human beings for insights. Yet those who have been seriously testing AI are reporting that it comes back with essentially average and uninteresting ideas.
Maybe those help someone else get some starting thoughts. But relying on AI leaves our own intelligence about design and creation unready for the true challenge of finding the exceptional — the things which return profits for companies.
In the book I’m writing, I note how the most serious problem facing innovation are the managerial structures of the companies themselves. AI will not fix that problem. Discovering products which matter — which make business strong — is not and never has been reliably predictable. Companies need to embrace that truth — not attempt to hide behind AI somehow magically making it “better.”
yes, Yes YES!!! Don’t think of Generative AI as a magic box that gives you the answer. Think of it as a team working with you to find answers, images, and designs you would not have hit upon alone. I use it as a data science colleague. It helps me refine math equations and write code in R or Python so I can implement. It does NOT replace me. But, it makes me better and gets me to a conclusion faster. It can work for any area of study. It also can be used as a tutor (Khan academy has created this) to function as a private tutor for any student, regardless of economic means. That will lead to a more enabled work force. Yes, Generative AI leads to better outcomes, but in a partnering and supportive way.
I absolutely believe generative AI will become a key tool in to product development. The amount of possible designs can explode by utilizing AI where a human just might not be able to think about all of the different possibilities. I think this also gives more power to smaller brands as well. However, there is still always the need for a stylistic, human eye.
“This same technology could (and I’d bet will) see application beyond fashion: in the film and games industries for generating concept art, advertising and media for generating content, architectural and aeronautical construction for concepts, and a myriad more. It’s a very powerful tool that the new age of designers and other creatives will have at their disposal to realize their visions and speed up their creative processes, much like Photoshop and its ilk have been in the past.
Where the article mentions that the AI doesn’t currently factor in material or practical aspects, this could absolutely be something it does very effectively in future iterations. Combining this GAI with an AI built for optimization, it could not only produce design concepts based on user-defined parameters, but also optimise these and group them by cost, material availability, physical capacity of machinery and cutting methods, and any other dimension to the search space that could be relevant. This would give the human designers and decision-makers more information on which designs would be best for production. Currently, it serves as an assistive tool for humans to use. Ultimately, it could get to a point where it can become self-autonomous, running as a process that humans can monitor at a higher level, without needing to do things such as vet individual designs.
In my opinion, there’s a delicate balance between AI and human creativity in design. While generative AI can certainly bring innovative ideas, human creators add the irreplaceable touch of emotion and intuition. While generative AI, as a supporting tool can accelerate the creative process it shouldn’t replace human decision-making. In fact, some aspects like the engineering and physical viability of a product are better handled by humans. It’s a symbiotic relationship where AI can enhance the efficiency in concept exploration, however, the final call on design completeness must stay with human designers. There is also a potential risk of overreliance, potentially dulling the unique human aspect of creativity. Ultimately, integrating generative AI may speed up development but a brand should prioritize maintaining the authenticity and quality that the human touch brings.
Generative AI will create designs faster than humans and more design detail will come from the AI processing. However, the final design for all products will come from a human review. Humans know better what the consumer will like and buy.