Look Forward gives you a glimpse of the Salon Traffic, the Fédération du prêt-à-porter féminin’s major event.
The closing conference, organised in partnership with Showroomprivé.com and moderated by Fabrice Jonas (Modelab), gathered Nicolas Latour (Digital University DDB), Antoine Millien (Fabernovel Innovate), Tony Pinville (Heuritech), Clarisse Reille (Défi), Alicia Birr (Google France), Guillaume Aurine (Salesforce) an Anil Benard-Dende, COO of Showroomprivé.com, around the question theme “when fashion questions AI”.
AI: a single definition?
First, we need to clarify what is implied with the term “AI”: it’s a cohesive set of processes that allow the automation of technical tasks. For instance, machine learning is on its declinations where machines learn to reproduce tasks after being shown an example. Multiple AI exists, and they are all specialized. We are still very far away from a Terminator-like machine that would resemble an actual human brain in all its complexity.
To date, AI is still limited to being a companion to the human brain: it can support it, complete and augment it, but it cannot replace it says Antoine Millien. On the contrary, AI could be the way to put humans back at the center of our processes.
AI and humans: the case of customer service.
Indeed, AI is already at work when it comes to automating tedious et repetitive tasks. Valuable time is thus freed for humans to work on subjects on which they have a real added value.
More precisely, in the case of customer service, AI can allow the machine to handle the automated aspect, and everything related to processes and information. The customer service customer can then focus on emotions, which are the core of the relationship with customers, says Anil Benard-Dende.
According to Guillaume Aurine, IA can free up to 20% of the human time by finding the right information and presenting it to humans at the right moment. This is already a reality; for instance, AI is what show us the right pictures when we do a search in our phones (“show me all my pictures of beaches”).
Although to Clarisse Reille, it is obvious that the emergency for the fashion industry is to focus on customer service, it’s also clear that new perspectives appear as all key players are now focusing on emotion.
AI and style
According to Anil Benard-Dende, “there’s no standard for product catalogs in fashion today”. This is a major dent in retailer/brand relationships. It is clear that AI will allow for major progress in this field, as more and more start-ups like Scalia seize the matter.
As for consumers, AI is already tangible. Product recommendations are already a part of everyone’s online journey. The next big step, says Alicia Birr, is that future AIs will be able to understand what style is. This change is already happening and will allow customers to not only get recommendations for similar products but also products that will complete a “look”. Such is the ambition of Amazon’s Echo Look for instance.
For designers, AI is becoming a tool to decrypt trends. Tony Pinville points out that although we are not yet able to predict what will happen tomorrow, we are already able to understand what’s happening today. It is now possible to feel the customer’s pulse, then inject it in the creative and research & development processes. AI can thus become a creative tool, allowing a finer understanding of consumer needs. And, says Nicolas Latour, “Personalization is the holy grail of every company”.
Can fashion survive AI?
The next challenge is that of human resources. Is AI going to destroy jobs in fashion, through the automation of part of the (so far) human tasks? There is no denying that some jobs will in fact disappear. However, new needs appear and new jobs will have to emerge. 60% of the jobs that will exist 10 years from now have now yet been invented: fashion will reinvent itself alongside the evolution of technology. According to Guillaume Aurine, the main challenge will be education, and how skills can meet industry needs. Investments need to be made today, so as to match the necessities of tomorrow.
You can also listen to the podcast of the conference (FR) : http://salontraffic.com/podcast_intelligenceartificielle