Artificial Intelligence (AI) has dominated the headlines recently, causing quite a stir. It has been powering tech solutions that rapidly transform industries, and the jewellery sector should be no exception. But while current AI tools can generate concepts quickly, they still require considered human expertise to turn those ideas into commercially practical pieces. Let’s delve into the current applications of machine learning within the jewellery industry, as well as where it might be headed:
AI-ASSISTED JEWELLERY DESIGN
So much software available to jewellers already uses AI algorithms to streamline the customisation and manufacturing processes through 3D renders and the simulation different materials and finishes. Beyond the familiar CAD design players, the application of artificial intelligence can be used in the jewellery trade for:
Sources of inspiration and fast-tracking the ideation process
All designers suffer from creative block at some point or another. Generative image applications can create a large number of designs quickly, based on specific parameters, so you don’t feel stuck staring at a blank page. They also may draw on influences you’ve not been exposed to, for example certain styles or cultures. These tools can serve as starting point for handcrafted jewellery designers, who then refine and curate the results based on their own preferences until an interesting concept is produced.
Here’s an article outlining one experiment which created images of jewellery inspired by the work of painter Piet Mondrian, and another which explores the results from multiple inputs. StarryAI is one free tool that claims to be able to assist jewellers with design, but with the rate that new ones keep appearing, it’s worth chatting to colleagues and keeping an eye out for the ones that work best for your industry. Stable Diffusion is probably a good option as it’s been around for a while.
For examples of how human designers are collaborating with AI to produce utilitarian items, see these Design Indaba articles on Spawn – the first spoon designed by artificial intelligence, and the Elbo chair. (pictured below, with an example of AI-generated ring design)
Prototyping
Image generators can also allow you to visualise what an existing design of yours might look like in different materials, variations or applications – for example from a ring to a necklace, or in different coloured stones or metals. Some image generators have the option of uploading an image to base their results on, which is handy if you’re tasked with creating an entire range from a single existing piece from your collection. Just like the normal 3D software many jewellers utilise, these prototypes can also be used for marketing the item before any physical handwork has been done on it, allowing you to pre-sell the concept before any costs have gone into the making of it – including your precious time – as these generative AI designs take considerably quicker to produce than any available sketching tools.
Selecting and working with diamonds and Gems:
- Gemtelligence is a platform that uses artificial intelligence to determine the country of origin of rubies, sapphires and emeralds, and heat treatments in rubies and sapphires. Unveiled in February 2023, it is a first in the gemstone industry, according to Gübelin Gem Lab.
- Ringo claims to be the world’s first AI diamond selection tool, able to identify diamonds that are “clean to the naked eye” for consumers based on their budget and preferred styles.
Generating pleasing arrangements of precious stones
Microsoft has investigated how machine learning tools could generate new aesthetically pleasing arrangements of gemstones (of all shapes and sizes), optimised for the space allocated for their setting. Leveraging jewellery design principles, the idea was to minimise the time jewellers need to spend on this phase during the design process. The outputs were then enhanced through a model called pruning, which discards designs with low visual appeal. When commercially available, this could very well save manufacturers working with heirloom jewels much time, and at the very least generate design options that provide a starting point from which their creativity and experience can be added. You can read more about this process here.
AI ASSISTED SALES AND MARKETING:
AI assisted e-commerce
Pioneering luxury retailers utilising AI in their online stores have reportedly seen sales increase by over 25%.
Digital try-on software uses an augmented reality (AR) interface similar to the filters available on apps like Snapchat and Facebook’s Spark AR to allow potential clients to see how a piece would look on themselves. James Allen developed one that also allows customisation of the item. This retailer also offers an online diamond inspection. Tiffany’s makes use of an app that can measure the size of a ring via a smartphone, as well as doubling as an advanced online catalogue. Creating custom solutions like this is mainly the privilege of the big name brands, but there are more accessible solutions emerging fast. Vossle is a cloud-based augmented reality platform that works on every modern smartphone browser, and requires no app installs. Their AR services include a virtual try-on function and there is a free plan. Many companies now also use chatbots on their websites to interface with customers either after hours or to free up their employees’ time.
AI for copywriting and content creation
Millions are already using AI to improve the fluency, clarity, and vocabulary of writing, to check grammar, or even build a free website (we can’t vouch for the quality of this one!). Here’s a video by jewellery business coach Megan Collins showing how you can use ChatGPT in your handmade jewellery business to write product captions, blog articles and social media posts, as well as email replies to unhappy customers.
CONCLUSION:
So will jewellery designers soon be replaced by AI? For the time being at least, the answer seems certainly not. As with most of the established tools (like Google search), the quality of the output is highly dependent on the quality of the inputs. A user’s expertise and creativity will influence the choice of parameters they feed into the model.
In addition, AI tools are not (yet) trained to be able to create designs that meet necessary manufacturing requirements and tolerances, or to adhere to industry standards for gemstone cutting and setting, let alone comfort in everyday wear, or the client’s budget.
Perhaps we should be embracing a future where AI will be fed a sequence of descriptors, generate a pleasing range of options and 3D print the chosen jewellery item without the involvement of any human hands. Embracing and becoming proficient in this new technology can optimise your workflow and save resources and time, which can be put towards continuous development of your own unique style, and to nurturing your client relationships.
In a not-so-distant world where an app could allow a layperson to create and make changes to jewellery designs in real-time, with no professional training whatsoever, your brand experience and personalised service will ensure you become the preferred option over a machine.
Trust will likely play a huge part in this process. Many consumers are already uncomfortable with algorithms that track their online browsing and purchasing history and preferences (which some developing AI tools make use of, in order to generate personalised designs for the general public). Plus, you will always find the kind of clients who don’t have time to do it themselves, and who value traditional techniques, craftsmanship and the artistic process that comes from human experience and feeling.
It’s likely that the average consumer will be unaware of the cutting-edge advances anyway. But to maximise your chances of success, you need to stay at the forefront of this emerging technology.