How retail professionals are adopting generative AI

The context

Whilst AI has been an integral part of retailers’ organisations for over 10 years, the deployment of generative AI marks a new stage. Since November 2022, when ChatGPT 3.5 was made available to the general public free of charge, adoption has been rapid and massive in the retail industry. This dynamic is reflected in the proliferation of new LLMs coming onto the market, both via proprietary and “open source” versions (Hugging Face).

Without delay, the major technology platforms have put this issue at the heart of their strategies. For the first time, Google is threatened, and forced to speed up new changes to search engines. Meta, meanwhile, has put the metaverse on hold and is redirecting 100% of investment into AI and LLM, which will be the biggest spend by 2023.

Use cases of generative IA are numerous. It is able to generate texts in all formats (online product pages, Web pages, summaries of customer reviews…) but also, with tools such as Midjourney or Dall-E, to generate images or visual backgrounds from a written prompt. In terms of Search, the tools understand what is overflowing in search intentions, by offering Web users a new use: a conversation with the machine, to better respond to fine-tuned and detailed queries. For example, “I’m looking for a vegan restaurant opening on 12 September near the Eiffel Tower, where I can bring my 3 and 5 years-old without too much of a wait, on Tuesdays at 1pm”. With natural language, consumers get a more precise answer than traditional search results via back-links. In fact, generative AI aggregates responses from several sources like company websites, Google reviews or social media comments into one single response.…