6.5 billion euros of investment in 2022. The retail world, like all sectors, invests enormously in artificial intelligence to optimize various of its processes such as logistics, collection, merchandising, customer experience…, it could be that new forms of so-called creative or generative AI without coding will take over. With them, more inspiration, more creativity and more creativity.
Create an image in a few words or the art of making prompts
Artificial intelligences such as Midjourney, Stable Diffusion or Dall-E Open Ai allow anyone to create unique and hyper-realistic images either based on an image or by simply typing a few words in a text box. The more precise and detailed the description, including for example the texture, color, size, style, etc., the closer the rendering will be to your initial request. This descriptive request is called a “request” or “prompt”.
These new applications are already very popular, particularly in the world of art or graphics. DALL-E 2 has over 1.5million users generating over two million pictures daily, while Midjourney has over three million members.
Proving the quality of the works created, a citizen of Colorado won an art contest by presenting a painting generated by MidJourney artificial intelligence. His victory is currently creating a controversy concerning creativity and image rights in the art world.
In view of the images below created by its AIs, beyond the hyper realistic and stunning side, an immeasurable field of possibilities opens up for the retail world!
Fast creativity for Fast Fashion?
If Chinese fast-fashion brand Shein uses AI to determine trends and predict patterns and changes in consumer demand, it could further accelerate the creation of new patterns with these new tools.
In the example below, my request was to create a bag loved by Generation Z and inspired by Rihanna, judge for yourself the result, especially since my description or ‘prompt’ is very sketchy. I used Stability AI (UK open source AI start-up for image, text and sound generation and has a $1 billion valuation!)
The Japanese company Datagrid goes even further by offering the real-time production of virtual and dynamic models in real time over the ‘prompts’ on e-commerce sites! a perfect complement for fast fashion? Are we heading towards an optimization of the ‘creativity supply chain’?
In this last example, starting from a photo of yourself or other, it is possible in real time to modify your outfits according to your most imaginative and trendy ‘prompts’ of the moment!
Finally if not, there are also models capable of generating illustrative images from the scan of a product. We can imagine the range of possibilities in advertising!
Beyond products, applications in concept store modeling
These new generative AI over time are able to design new products but also shopping centers or even concept stores by integrating the charter of manufacturing constraints while basing themselves on the style of a great architect! We can thus discover online, as examples, new Ikea store concepts and instructions for use inspired by Escher, or new store concepts for McDonald’s!
It is even possible to mix these different applications to have a realistic video rendering. Here it is the integration of a project made on slab-E in @runwayapp to have a simulation of the rendering in video and in a context of reality (here the example for the simulation of a future Chinese restaurant).
Starting from a book of charters including conceptual ideas, more objective architectural constraints, even the optimization of customer flows, it will be possible via a written transcription to have a realistic content of what the future architectural project will be.
Redesigning your interior will be done in a few words and a photo. With interior.ai, a photo of your interior, plus a description written by yourself or a choice of a theme proposed by the AI and you can have your dream interior simply by expressing your ‘desire’! Imagine the impact for specialized distribution in the years to come!
GAFAM on the lookout
If today these different applications are fragmented in the AI landscape, the tech giants are already starting to want to integrate them into their consumer and especially professional software such as the Office suite! Microsoft announced the arrival of AI based on DALL-E In 2019 (Microsoft invested a billion dollars in OpenAI, the firm behind DALL-E) in its 365 (Microsoft Designer) suite. This should greatly democratize the use of this technology.
Generative AI enthusiasts predict that this technology will take hold in all kinds of industries. David Song, a Stanford University student who follows the evolution of this phenomenon, compiled a list of more than 100 generative AI start-ups. They work on applications such as music generation, game development, writing assistants, customer service robots, coding aids, video editing techniques, VR…
There are already platforms, marketplaces that offer to put you in touch with ‘prompters’ each specialized in their prose and subjects like the public writers of a certain era!
We are only at the beginning of generative AI, but also of endless related ethical and legal debates (for example Getty image refuses any image generated by these apps, what about copyright?). Finally, generative AI is already in the Metaverse, a virtual world, where you can change your ‘virtual environment’ by voice!
About the Author :
With a master’s degree in distribution management in hand, Guillaume Rio joined the Auchan Group on several “digital store” and “merchandising” topics, particularly for Norauto structures and Auchan hypermarkets. With more than 20 years of recognized expertise in deciphering developments in commerce and international forecasting, Guillaume Rio now leads the Techno Trends Department of Echangeur by BNP Paribas Personal Finance.