The jewelry world knows that Artificial Intelligence will revolutionize its businesses and practices. But how and to what extent will it impact them?

By Sandrine Merle.

 

 

When we think of Artificial Intelligence in jewelry, we immediately think of creation and the inevitable disappearance of designers, art directors, gouache artists and artisans. As the excellent study by Comité Colbert and Bain & Company (link to the study at the bottom of the article) explains, the revolution has already begun in production and logistics at international jewelers such as Cartier, Tiffany&Co. and Louis Vuitton. With hundreds of points of sale worldwide and millions of data points on stock flows, customer tastes and local events, in-house data scientists and engineers have been able to create their own AI ex nihilo. This makes it possible, for example, to adjust the quantities of Alhambra in mother-of-pearl or turquoise, of “Love” in yellow or white gold, etc. according to the store and the tastes of customers from Los Angeles to Tokyo via Paris. Without eliminating chance and approximation, AI also enables us to reduce the impact of, and prepare responses to, unforeseen events likely to disrupt the supply chain: changes in customs regulations, variations in import taxes, climatic disasters, etc. It deploys the best scenario near the customer. It deploys the best scenario in almost a quarter of a second.

 

Avatar and digital twin

AI doesn’t just manage the flow of goods; it can also improve customer relations. We’re not there yet, but the Comité Colbert study predicts that before long, customers will be chatting with their sales consultant’s avatar. This avatar will not only present the new collection, but will also answer any questions she may have, advising her on the right jewelry combinations, or on the ring she has just inherited from her grandmother. In a matter of seconds, he’ll tell her the exact year of the jewel, the name of the collection, the source of inspiration and even the sketches made by the designer at the time, or the advertisements in which he appeared. It can even help by suggesting a resale price! This is because the AI has access to all the data stored by the company, which not even the most efficient sales consultant can memorize. For her part, the customer could very soon have her digital twin, i.e. her virtual representation with the same tastes, the same operating mode, the same behaviors reproduced thanks to a history of saved data (orders, chats, etc.). For the time being, it’s hard to imagine that the sale of fine jewelry, characterized by pieces flirting with the million-euro mark or more, could do without the sacrosanct physical relationship with the big seller… But why not?

 

What role can AI play on the shop floor?

Until now, workshops have been dominated by human labor. Frame-making, gold polishing and stone-setting are still carried out by specialized craftsmen. Even CAD (Computer-Aided Design) has not yet been taken over by AI. The two are often confused: Artificial Intelligence is capable of simulating human cognitive processes such as learning, reasoning and decision-making, while CAD is a modeling tool. AI can enrich CAD, saving time on complex techniques or repetitive tasks such as labeling the thousands of elements making up a piece of jewelry. But it is not yet capable of creating a structure, in other words, a skeleton with precise dimensions, locations, articulations, etc. All this still needs to be done by a human. But it’s only a matter of time: we expect to have this perfected in two- or three-years’ time,” says Valérie Leblond, co-founder of the virtual studio BLNG, which won the LVMH prize at the last VivaTech trade show. Grégoire Talon, coordinator of the Conservatoire des Gestes de Métiers d’Art et de Fabrication, insists on the positive side of AI: “It is already being used to document practices and train new craftsmen. And tomorrow, the time savings it will generate will help maintain jobs in countries with high wage costs. Every jeweler, every setter, every polisher will also have his or her digital double, a mentor capable of correcting his or her gestures, which, in terms of health, will minimize the risk of musculoskeletal disorders, for example.” This opens infinite possibilities for jewelry workshops.

 

Creation: the fantasy of AI’s great replacement

In the field of creation, AI generates phenomenal fears and reticence, the same fears and reticence that resurface with every new disruptive innovation, as they did in the early 2000s with CAD… This time, caution is more than ever the order of the day, because we know that AI is capable of multiplying inspiration from a blank page in 20 seconds, and translating a creative intention into a piece of jewelry. The very opposite of the values of haute joaillerie, whose marketing strategies extol the virtues of handcrafting, time and authenticity! The risk is to obtain standardized jewelry, “lukewarm porridge, because AI only feeds on what already exists”, asserts Grégoire Talon. Even if we’ll probably get used to this style and soon consider it as the trend of the 2030s…”. In any case, it’s not enough to prompter make-me-a-pretty-engagement-ring: you must know how to express the thoughts and visions that it must translate with an elaborate and relevant vocabulary. You also need a sense of color and proportion… Everyone can use Word, but not to write a good text,” reassures Ludovic Blanquer, Creative Director of Francéclat. The big revolution to come is the possibility of hyper-customizing jewelry thanks to virtual studios like BLNG, to the extent that tomorrow, Victoire de Castellane, Claire Choisne and others will co-create special orders directly with their customers (and the marketing department), saving weeks of exchanges. “The customer is thus engaged in a very satisfying experience”, confirms Amine Messaoudi, who is already testing this type of tool for Atelier Mille Or.

 

Tomorrow, a jewelry Shazam?

Thanks to AI, we may soon be able to count on a kind of jewelry Shazam… 58Facettes, a platform for second-hand jewelry, has developed an AI capable of identifying a piece of jewelry from a photo and information provided by the owner (gold grade, type of stone…). It then provides an estimate. “We’ve developed it on the basis of more than a million results, notably from auction sites, and we’re constantly adding to it,” explains Alexis Blez, co-founder of the platform.  There’s no doubt that the human expert, who sees thousands of pieces of jewelry in his or her lifetime, can no longer compete… But today, his or her eye is still needed to check that the machine isn’t mistaken, because while it can estimate an ultra-famous piece of designer jewelry such as a Cartier “Clou” bracelet or an “Alhambra” necklace (and provided that the information given by its owner is accurate), it is not yet capable of distinguishing between, for example, a natural stone and a synthetic stone or glass. “If we train it on x10 photos of perfect quality, we can imagine it achieving this in the not-too-distant future”, says Alexis Blez. Nor has she yet mastered the subtleties of unsigned jewelry. Recently, it valued a Suzanne Belperron brooch at 940 euros when it was worth at least 45,000. Don’t trust AI blindly, and keep a critical eye: don’t panic, humans still have their role to play.

 

Banner image © BLNG

 

Download the Comité Colbert / Bain & Company study