Artificial intelligence (AI) is shaking up the foundations of ecommerce, and opening up new opportunities for years to come, says Mauricio Monico, Chief Product Officer at Wish in this guest post today:
ChatGPT and, more broadly, new forms of AI such as Generative AI are currently the subject of much debate, with questions and concerns being raised around the potential impact (both positive and negative) of these new technologies on our society as a whole. While the advent of AI brings with it opportunities to innovate at an accelerated pace, there are real risks associated with its widespread deployment, which need to be managed carefully.
While artificial intelligence is not new, it is the proliferation of AI into our increasingly digital and connected lives that is inviting a heightened level of scrutiny. If we take the example of ecommerce, where AI tools are already ubiquitous, we can see clearly how it is being used as a force for good.
AI-based tools have been widely used in ecommerce for many years. Artificial Intelligence forms the backbone of many technologies being used to facilitate user searches, personalise a user’s product feed, push recommendations, advertisements and much more. We need only look at the deployment of AI-powered chatbots in the pre and post-purchase experience to see how AI has, in many ways, become the norm.
More recently, the CLIP (Contrastive Language-Image Pre-Training) model developed by OpenAI is being used to further the understanding of products, and improve classification and merchandising practices. It involves extracting the relationship between images and their corresponding text (e.g. titles, descriptions, keywords) and creating tags that describe product features, themes and moods, and leveraging this understanding to surface the most relevant products to each user, in a highly personalised way.
Similarly, by leveraging such technology, it is possible to generate theme, interest or attribute based tags and collections of products for users based on their browsing and purchase history. This allows ecommerce platforms like ours to better understand users and shoppers’ preferences, and surface products they are likely to engage with or purchase. This is an essential tool for developing discovery and interest-based shopping experiences that satiate the needs of the masses – from film buffs to kung fu enthusiasts.
In recent weeks, Generative AI has been making headlines as more and more companies come forward and talk openly about their AI use today, and share their projections for the future. For Wish, Generative AI presents a huge opportunity to remove friction from the shopping experience. It helps us guide users on their shopping journey by addressing broader requests such as “build me a basket of engagement party decorations for less than 20 euros”, and presenting similar or complementary content based on the products a customer is (or has been) browsing. Through Generative AI, we can incorporate content-based signals into our search and personalisation systems (that already have signals from user interactions) to improve recommendations throughout our platform. In addition to recommendations, it has allowed us to present titles that more clearly represent the product being sold to improve a user’s confidence in their purchase decision.
Today, technologies already allow ecommerce platforms like Wish to identify the most popular products, present them to consumers, better manage trends and, consequently, stocks. Tomorrow, the ability to accurately predict demand will enable merchants and manufacturers to more efficiently manage their supply chains, and logistics networks.
But we are only really scratching the surface. As AI tools become more sophisticated, so too will our demands and expectations.
AI represents a major opportunity to make ecommerce more virtuous. By leveraging AI’s analytical and predictive capabilities, we can build systems that allow us to reduce waste, optimise inventory management and promote sustainable practices. By working together and aligning our desires with those of civil society and the regulatory framework, we can use AI as a powerful lever to shape a more responsible ecommerce future.
