LONDON, United Kingdom — If artificial intelligence has its way, discounting could disappear, thanks to software that tells retailers exactly what and how many products to buy, and when to put them on sale to sell them at full price. Online shopping could become a conversation, where the shopper describes the dress of their dreams, and, in seconds, an AI-powered search engine tracks down the closest match. Designers, merchandisers and buyers could all work alongside AI, to predict what customers want to wear, before they even know themselves.
In the last few years, a trifecta of cheap, ubiquitous, powerful computing; big data; and the development of deep learning have triggered a revolution in artificial intelligence. The computing devices that now fill our everyday lives generate large data sets, which “deep learning” algorithms analyse to find trends, make predictions and perform specific tasks, such as identifying specific objects in an image. The more data presented to the algorithm, the more it “learns” to do a task effectively.
Earlier this year, in a blog post titled What’s Next in Computing?, Chris Dixon, partner at the venture capital firm Andreessen Horowitz, wrote, “Many of the papers, data sets, and software tools related to deep learning have been open sourced. This has had a democratising effect, allowing individuals and small organisations to build powerful applications.” As a result, AI might “finally be entering a golden age,” he wrote.
No area of life or business will be insulated from AI, in the same way that no part of society hasn't been touched by the Internet.
These developments have provoked an AI arms race. Companies like Google and Apple are snapping up AI start-ups, and in the last year, milestones in the field have arrived faster than previously expected, such as last month, when Google's AlphaGo program beat a human champion at Go, a strategy board game considered more complex than chess.
Already, big businesses are using AI — Kensho, a data-crunching AI software, is automating finance jobs at Goldman Sachs, while Forbes uses AI to automate basic financial news stories. IBM’s Watson — a set of algorithms and software that is the company’s core AI product — is available as a cloud service, enabling research teams to rapidly analyse large amounts of data, such as millions of scientific papers, to test hypotheses and discover patterns. By 2020, the market for machine learning applications will reach $40 billion, according to International Data Corporation, a marketing firm specialising in information technology.
“No area of life or business will be insulated from AI, in the same way that there's no part of society that hasn't been touched by computers or the Internet,” Kenneth Cukier, data editor at The Economist and author of books including Big Data: A Revolution that Will Transform How We Work, Live and Think, told BoF. “Today it seems shocking because it's new. But in time, AI will fade into the background as just the way things are done.”
By presenting a cheaper, faster way of doing many tasks that companies currently employ humans to do, many predict AI will radically alter industries from transportation, to healthcare, to finance. In fashion, like in other industries, driverless trucks will likely reduce companies’ logistics costs, or software like that used by Forbes could be used to write formulaic text, such as product descriptions on e-commerce sites.
But for fashion, some of the biggest opportunities are in aligning supply and demand, scaling personal customer service, and assisting designers.
Aligning supply and demand
Currently, fashion brands and retailers work with a limited amount of data, to predict what products to order and when to discount or replenish them. If they predict wrong, the result is loss of income due to mark-downs, waste and popular items selling out. By analysing large amounts of data — say, the browsing and shopping history of every single one of a fashion brand’s online customers, as well as those of its competitors — AI can tell a retailer how to align product drops to match demand, and even how to display products in a store to sell as many as possible.
AI’s ability to make predictions like these has particular implications for a trend-driven industry like fashion. Today, the fashion market is visible online: an AI can crawl e-commerce sites to see which products are selling; it can analyse consumer data to learn which colours or materials customers in a specific country — or even city — are buying; and it can scoop up swathes of information from social media to identify trends and microtrends. This data — which was not previously available — could help brands be first to market with styles that are likely to become mainstream trends.
Edited, a data analytics company specialising in fashion, is already doing this. Edited's software has “learned” to recognise apparel products in images, and natural language processing software, which can classify these products. Edited let this loose on a bank of data on 60 million fashion products, collected from retailers and brands in over 30 countries, in over 35 languages: the result is a searchable database of organised, structured information on each of these products.
“We can process the data in seconds. No one could ever do it manually,” says Geoff Watts, chief executive officer of the company. Brands that work with Edited “usually start by analysing their competitors’ historical pricing and assortment data to make more strategic decisions, ultimately leading to better sales, stronger inventory management and less discounting,” he says.
Ganesh Subramanian, former chief operating officer of e-commerce giant Myntra, and now co-founder of Stylumia, an AI-powered tool for fashion professionals, agrees that AI could stop fashion companies making important decisions in the dark. “A trend is nothing but a movement which has a beginning and a gradual adoption,” he says. Like Edited, Stylumia uses AI to make sense of a sea of data, from videos, e-commerce sites, social media, etc. “We can not only spot trends, but also come out with what is the relevant timing for [brands and retailers] to adopt,” he says.
Scaling personal service
In the days when luxury goods could only be bought in a few physical boutiques, one-to-one customer service was at the core of the industry. The Internet changed that dramatically, giving customers a seamless — but often impersonal — way to trawl thousands of products and purchase without exchanging a word. Could AI deliver that original one-to-one service at scale?
One way to do this is through chat bots, which can exchange messages, stories and information with humans. Already, Microsoft’s XiaoIce chatbot is being used by 40 million people on Chinese microblogging platform, Weibo. (Not all attempts to have bots interact with humans have been so successful: when Micosoft unleashed Tay, another chat bot, on Twitter last month, the bot “learned” from other users and rapidly began tweeting offensive messages.)
Machine learning can also enable brands to finely personalise their offerings to each market, or even, each individual customer. Thread, an online personal styling service, combines human stylists with machine learning algorithms. The AI crunches data like what human stylists thinks would suit an individual user, where they live and what the weather is like there, as well as the user’s ratings of products on the app, which items they click, and how customers with similar purchasing habits responded to product recommendations. The AI then trawls through 200,000 fashion products and makes a judgement on what products to recommend.
“Humans are limited in many ways,” says Thread founder and chief executive officer, Kieran O'Neill. Not only can AI process a vast amount of data — it can also “remember your preferences in a way that it’s just not practical for humans to do. A computer remembers everything,” he says. Michele Goetz, principal analyst covering cognitive computing and data at Forrester, agrees: “That’s where I think AI shines, being able to scale insight.”
IBM's Watson — which is working with over 500 partners in industries including retail — has partnered with The North Face to offer “guided shopping” online. The AI asks shoppers questions on factors such as gender, time of year and technical product details, to deliver tailored recommendations. "Online shopping can be overwhelming. There are so many choices and products from so many different sources,” says Keith Mercier, ecosystem manager of Watson. AI, he says, “can help retailers make sense of massive amounts of unstructured data to improve and personalise the online shopping experience."
Image recognition apps such as Snap Fashion and ASAP54 are also harnessing AI to build search engines for fashion. In theory, a user can snap a picture of someone on the street wearing a dress they like, or even something as abstract as a painting, and an image-recognition search engine will search a huge database of shoppable products and serve up similar items. When BoF tested these products, the search results were far from perfect, but Kieran O'Neill bets that “in the next three years it will become pretty good.”
"There are AI systems today that compose music, write stories, and create artwork that no one can tell is machine-generated. So fashion design is surely not beyond AI's capabilities,” says Pedro Domingos, author of The Master Algorithm, which predicts the revolutionary impact of machine learning, “What will likely happen, however, is not that AI will completely replace designers, but will become an indispensable tool for them."
In the same way that the work of architects like Frank Gehry and Zaha Hadid relies on computer modelling, “Fashion designers armed with AIs will be similarly able to come up with radical new ideas: AI will amplify their creativity rather than replace it," reasons Domingos.
“AI will absolutely challenge and replace designers,” counters Kenneth Cukier. “Let's get real — lots of design is trial and error or boring, repetitive work. AI can help with both by making more accurate predictions of what designs will work and taking over some of the repetitive tasks.”
Approaching AI now
Some believe fashion brands should strike early and invest. “They certainly need to have in-house AI teams, like other companies, whether by building them from scratch or by acquiring start-ups,” advises Domingos. “Those who wait and see risk falling behind, particularly in a fast-moving industry like fashion, where consumers are the main drivers and tastes are fickle.”
“The old world of personal touch is not necessarily going away, but it’s not the way you’re going to grow your brand even from a luxury standpoint,” argues Michele Forrester. When fashion brands thing about AI, she says, they need to consider the next generation of luxury customers, who were born into a world of social media, and handed at birth the ability to buy anything they want, from anywhere in the world. “They don’t have the patience for a one-on-one relationship,” she says.
Indeed, the next generation of big spenders is already using AI: GPS navigation shapes their driving habits, while algorithm-driven personalised recommendations from Spotify and Netflix influence the songs and shows they consume. “If you don’t have it, you are not aligned with the experiences they’re used to,” warns Goetz.
Others are more cautious. “The top tier brands should resist the temptation to buy into the AI world right now,” says Cukier. “Their business is being good at fashion, not smart at technology… Right now, the most promising technologies are still in the lab or in field trials, like self-driving cars. Big, smart, non-technology companies can afford to wait.”
Others agree that, for the moment, partnering with third party AI specialists is the way forward. “The smartest thing a business can do, is partner with a fashion-focused tech company with AI at its core,” says Geoff Watts of Edited. “Building AI teams from scratch, or acquiring AI start-ups and retrofitting them to have a retail focus, requires a substantial investment of time and money.”
Kieran O’Neill of Thread adds that, rather than dive straight in to AI investment, brands should build a strategy around AI, and work out on what the lowest hanging fruits are for their business. Some of the brands using Thread — such as Burberry, Jigsaw and Topman — signed up to sell on the platform, not because they needed the sales, but “because they really want to be close to the AI stuff we’re doing,” he says.
“Every company in every industry should be paying very close attention to AI,” advises Martin Ford. “There is no limit to how far it can go.”