Artificial intelligence, otherwise known as AI, has arguably been the most disruptive force of the last decade, upending almost every industry, from manufacturing to healthcare, agriculture, banking, cybersecurity and music. But when it comes to the more creative industries, art, TV & radio, crafts, design and fashion among them, we are yet to understand what AI’s full impact will be. Fashion – a $2.4 trillion industry – has so far been surprisingly unaffected by the technology, but there is no denying it is ripe for the picking when it comes to disruption.
Given that AI is a technology able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, and decision-making, it’s actually shocking that AI hasn’t yet been adopted in a big way to advise individuals and designers on their fashion choices. There are so many ways in which AI could make a massive impact with respect to fashion. Consider this: through capturing images of individuals bodies and then searching the internet instantaneously for suitable clothing, makeup and accessories, it would save shoppers thousands of hours a year spent browsing, deliberating and wasting otherwise valuable time searching for products that suit their physique, style, colouring etc. I personally would spend roughly triple the amount I currently do on female fashion were it an easier, less painstaking task. And I cannot count the number of times I wished there were an app or a device that could simply tell me what to wear in the morning, or what type of dress suits my figure best.
Amazon, Google and Facebook have all begun to use machine learning to give users tips on what to wear, so are we soon to see the technology roll out industry-wide? Already, Google Image Search is capable of helping shoppers find similar products online by inviting web users to drop an image of an outfit and then using machine learning to match patterns and common features between the image and the global database. Google Lens and Amazon Style Snap can also tell you more information about the garment, including how it should be worn as well as where it can be bought, while Amazon Echo Look is a voice-controlled camera that works as a style assistant, comparing two photos using a machine learning algorithm to tell you which looks best. Though it has received subpar reviews so far, it shows promise for a future where difficult daily style decisions can be palmed off to a device, saving both time and stress. Facebook’s Fashion++ project is a more advanced version of a style assistant: using a deep image generation neural network, it tells you how to improve the outfit you’re wearing by suggesting which accessories to pair your clothes with, whether you should tuck in a shirt or not, or whether you should remove those clip in hair extensions. Small, practical changes, but ones that can have a meaningful impact on a person’s style and look.
Take AI a step further, and one day it could actually replace designers. Using machine learning and data to predict high-probability best-selling garments, the manufacturing end of the fashion industry could run entirely off AI technology, with products being on the shop floor as soon as four weeks after the design has been made – an improvement from the average 18 – 24 months it takes today to end up on the shelf. AI technology could pair with other automated technologies to entirely remove the design process from human hands by spitting out a whole set of machine-generated designs, as well as liaise with logistics department and manufacturing teams to order the material and create the clothing. Could you imagine? Removing the ego, fervour and time-wasting habits of fashion designers would be blessing to all those working in fashion, and no doubt the clothes would be much cheaper, too.
Better yet, the results of clothes products via this process are incredible and there is far less waste at the end of it (both in terms of material, and clothes left unsold at the end of the season). Machine-generated designs are simply better, usually selling at two to three times the regular rate of sale.
There are a ton of other potential applications for AI within the fashion industry too: wearable solutions, enhanced customer service and communication, improved logistics, easier inventory management, and so forth. The best part is, there are really no ethical concerns at play when it comes to utilising AI for individual and commercial fashion advice, unlike many other applications of the technology.
For an industry that has been around for over 200 years and which has changed very little in that time, it goes without saying that fashion is ready for major disruption. Will AI be that disruption? Only time will tell.