In 2018, an artist named Robbie Barratt used artificial intelligence to come up with an entire Balenciaga-inspired fashion collection. Barrat had to train his own AI model based on images of existing looks he had collected himself, but the output wasn’t perfect. Many of them looked dirty and warped. But it was enough to catch the attention of the fashion industry, resulting in an interview with Ssense and his collaboration with Acne Studios.
Just a few years from now, anyone with a computer will be able to use new generative AI tools to essentially create a collection with less effort than Barrat’s previous attempts, with more photorealistic results. became.
Generative AI, which writes machine learning algorithms that can create new content, will have a major impact on fashion brands. Trained on enough examples, these algorithms recognize underlying patterns and structures in the data and create new examples of their own. Some tools such as DALL-E 2, Midjourney, Stable Diffusion can generate images. Some, such as ChatGPT, can generate text. Some, such as Runway, can also create videos. However, both can create content from text prompts, so they can be widely used.
Generative AI has gained prominence online with a flood of user-generated memes, such as videos of Balenciaga-esque reimaginings of the Harry Potter films, but its biggest impact may be on business. Tech giants such as Microsoft and Google vie to incorporate it into their core offerings, while industries ranging from healthcare to finance seek ways to harness it to boost labor productivity, and many start-ups are on board. are building enterprise applications. Analysts at Goldman Sachs believe that broad adoption of generative AI could boost U.S. productivity by 1.5 percentage points annually over a decade, ultimately boosting S&P 500 earnings by more than 30 percent over that period. We estimate that it is possible to
Fashion will experience its own growth influenced by generative AI. While the industry is only just beginning to explore the technology’s potential, McKinsey & Company has the potential to add $150 billion to $275 billion to its apparel, fashion and luxury sector operating profits over the next three to five years. I estimate there is. It’s already starting to make a big impact in areas like design, marketing and customer service.
As technology is being tested and deployed in different parts of the fashion industry value chain, this case study explores the opportunities, challenges and risks for both brands and their employees. However, the field is changing rapidly, and the use case was a work in progress when we conducted our research and interviews in the spring of 2023. That said, his four key strategic areas have already been revealed when it comes to fashion, so this case study provides a glimpse into the current state of the fashion industry. Product design, visual content creation, copywriting, shopping and customer experience. Join brands and retailers such as Revolve, Casablanca Paris, Frame, Snipes, Levi’s and Zalando to help build the knowledge provided in this case study. Technology providers such as Salesforce, Amazon, Shopify. AI-focused experts such as will share real-world examples of leveraging this emerging technology and what they’ve learned along the way.
Glossary
Artificial Intelligence (AI): Technologies such as computer programs that mimic the human brain’s ability to perform tasks and learn and improve over time.
Deep learning: A form of neural network involving three or more layers of nodes. These algorithms are more effective at processing and understanding unstructured data.
Popular model: A type of generative model trained by adding “noise” to a set of training data and reversing the process to reconstruct the data. Random noise can be used to create new data by repeating the “denoising” process.
Generative Adversarial Network (GAN): A type of generative model that can generate realistic images using two neural networks, a generator and a discriminator. Generators create new content and discriminators try to distinguish whether it is real or fake. As each improves, it becomes increasingly difficult to identify that the output is false.
Generative AI: A deep learning algorithm that can use enough examples of something to recognize underlying patterns and structures in data and generate new iterations containing text, images, videos, etc.
Machine learning (ML): A branch of AI that includes algorithms that can identify patterns in data and learn from them without being explicitly programmed.
neural network: A form of ML that uses a series of interconnected algorithms structured in different node layers that mimic how the human brain works.
Transformer: A neural network that uses sequences of data, such as words in a sentence, to identify relationships between data points that may be far apart from each other, allowing it to understand context and predict new sequences. Many language models, like OpenAI’s GPT model (GPT stands for “Generative Pre-trained Transformer”), are transformers trained on huge amounts of data.
