Artificial Intelligence is already impacting the future of everyone and everything in ways we can’t yet quite comprehend, and the watch world is no different. AI programmes are already used in various software across the watch industry to assist with numerous tasks, but with the new technological dawn brought about by the likes of Chat GPT and Text to Image AI, tools such as DALL-E 2, and Stable Diffusion, the AI watch design landscape could shift entirely.
As someone excited by cutting edge tech while also fearing it, it’s truly a unique time to be alive. Despite what you might read about the watch industry being slow to catch on to trends, the past ten years has seen brands dabble with blockchain tech, virtual reality, NFTs, e-sports and more. The next heir to this technological throne is surely AI and while early tech is rife with scams and gimmicks (I’m looking at you NFTs) there’s no doubt they offer some great use cases.
At the time of writing, the three front runners in the image generating software game are DALL-E, Midjourney and Stable Diffusion. DALL-E was created by Open AI, the same non-profit behind the text predicting ChatGP, while Stable Diffusion and MidJourney operate on their own separate software. For reference, Midjourney 5.1 has been used to generate all the AI watch design images in this article.
Using Midjourney is incredibly simple. Via the Discord sever (check your app store) the user enters in text based ‘prompts’ and the AI spits out images with incredible proficiency, using tens of billions of data points as a reference. The results are often creative, succinct, surprising and (if you didn’t know it was generated by 0s and 1s) imaginative.
While it is incredibly easy to get started, the more you understand and articulate both artistic and design language, the easier it will be to achieve the desired result. This is something I struggled with initially, however there are corners of Discord dedicated to ‘prompt chat’ in which people give guidance on specific terms, such as ‘realism’, ‘photorealistic’ and ‘aspect ratios’, or broader ideas that have worked for them.
As well as the /imagine function (the one you need to put in to conjure up the images), there is also the /describe function, which will essentially describe any image that you input. This function is really useful to mimic existing imagery or to help understand the language behind images according to the AI. It’s why you see some of my prompts include things such as Royal Oak, or Patek Phillipe and you can see examples of the prompts used in the image captions.
Despite achieving some pretty cool AI watch designs, some of which I would love to see made, the jobs of watch designers (or any other designers for that matter) aren’t in any imminent danger. My experimentation has shown that the ideas the AI generates are brilliant conceptualisations, but it’s extremely hard to achieve absolutely what you want; something that can of course be done by a skilled draughtsman. In addition, the images also lack detail and practicality. In other words, they’re great concept designs, but not so great final designs.
So instead of being replaced by AI, Midjourney is simply another tool in the designers’ arsenal. The future Gentas and Hyseks will use AI symbiotically, however it will no doubt be used by nimble and innovative brands, as opposed to the giants of the industry.
Of course, this is just the beginning, with each version AI software is making huge leaps, and no doubt more real- world 3D design applications will soon be released. This would mean designers could account for specific details such as sizing (both externally and for housing specific movements), materials and integrating existing components.
Regardless of your opinion on AI, and whether or not it’s going to destroy us it’s here to stay and for better or worse its making design more accessible – and I’m all for it.