Will AI-Centric Devices And Agents Reframe How We Interact With Retail?

Sign up to Interline Insiders to receive our weekly analysis in your inbox every Friday.

Next Week: Munich Fabric Start Sets Out To Help Fashion Find Clarity

In the face of what feels like perpetual uncertainty everywhere from raw materials and logistics to trend directions and consumer appetites, fashion businesses are still finding it difficult to put a confident foot forward. So whether it’s to fuel pricing, creative direction, fabric choice, assortment building, merchandising, or marketing, clarity and certainty remain the industry’s most sought-after currencies.

Taking place next week, Munich Fabric Start aims to bring the conversations, connections, inspirations and innovations needed to source that clarity to life. Find out more about the themes and the principles behind next week’s show here, or explore a full agenda and keep up to date on future events via the Munich Fabric Start website.

Are We Headed For An Era Of AI Agents? And What Could It Mean For Retail?

Surprising nobody, AI is proving to be the most fashionable investment of 2024 – at least so far. Brands and retailers across the board are making deep investments in creating their own AI frontend experiences and AI-enabled applications. Either in off-the-shelf ChatGPT form or through more bespoke models, AI is being used for everything from answering customer service inquiries to inspiring new collections. 

Quite how the use of generative image models for design will shake out is, right now, anyone’s guess. This remains a tightrope, with copyright infringement and job losses on one side, and efficiency and a potential redefinition of creativity on the other. As evidence by the overwhelming presence of AI at the NRF Big Show, though, the through-line from the precipitous rise of ChatGPT to new, AI-powered retail experiences is just getting started. Modern Retail reports: “More than 20 exhibitors have “AI” in their titles, while companies such as Google, Yoobic, and Salesforce released new AI tools or research in time for the show. There were more than a dozen events about AI on Sunday alone, according to an agenda posted online.”

This comes as no surprise, as some of the biggest retailers – and global powerhouses – have already gone all-in on AI. Target has offered customers chatbots for a long time and is continuously working on improving them. Walmart recently announced two new AI-powered tools to help shoppers quickly search for products and automate the process of re-ordering frequently ordered items. And Amazon is now fully on the AI train, investing in AI-enabled personalised size recommendations, review highlights, re-imagined size charts, fit insights, and now its own in-app chatbot. Generative AI is also being used to produce text and visuals for marketing initiatives, as well as refining product descriptions to enhance search optimisation.

All of this is in service of streamlining and optimising the retail experience. Companies are making significant bets, as a consequence, on the idea that shoppers will continue to come to their retail channels for that streamlining to take effect. But what if AI ended up doing something more fundamental to our interactions with websites, apps, and services that made it less likely that consumers would land on those retail channels to begin with?

We’ve already seen – and written about – the changes in product discovery and online shopping that are stemming from Google’s experiments with gatekeeping eCommerce, and about the shifts taking place at the heart of search that are leading to an internet of content created by robots, for robots. But even those ambitions were predicated on the idea that eventually, after the AI models had made their recommendations, the entity doing the purchasing would be a person.

But over the last week we’ve begun to see suggestions that what people might actually want – or at least what technology companies are gearing up to sell them – is a different model: one where AI agents act on our behalfs, using apps, making purchases, and much more. 

This week, for example, Samsung Electronics hosted Unpacked in San Jose, California, where they unveiled an “all-new mobile experience powered by AI.” In a press release, Samsung boldly called their newest mobile experience “revolutionary” and stated that customers should prepare to have their lives altered for the better. “Get ready to discover a new era full of possibilities with the latest Galaxy innovations, designed to transform how you live, connect and create.” 

In practice… what the consumer electronics giant delivered was a regular smartphone with some added AI features, but that reveal event came soon after an AI startup Rabbit announced that it sold through the first 10,000-unit production launch of its standalone AI device, called the R1, in just one day. That device, if its lofty promise ends up being realised, is much closer to the vision for a genuine AI assistant: a set of action-oriented models that are expressly trained to operate apps, websites, and services on behalf of their users.

If this agentic future plays out, it could have a huge impact for fashion retail: if we stop using our phones as we currently do – no more tapping around on apps and doing online shopping, but instead asking AI agents to do it for us – our experience with e-commerce will be completely different. We won’t need brand-specific chatbots or single-retailer AI-assistants, this role will be filled by our service-agnostic AI agents.

While both are getting considerable buzz, there’s also the growing belief that on-device AI represents the next leap in mobile computing, and this is almost certainly the tipping point that Apple et al are waiting for before launching their own first-party AI assistants. Very few people would argue with the idea that Siri needs a radical overhaul, and on top of the privacy arguments, running compact models directly on our devices also enables faster processing of requests. And by bypassing the need to rely on the cloud, we could be able to instruct AI for tasks that typically involve multiple follow-up requests and that require some measure of knowledge and interaction with what’s displayed on our screens, and with the apps and services we are signed into.

This seems like a step in the right direction – if only because it stands the chance of realising the decade-old idea that digital assistants would do our bidding for us across our personal and professional levels. But even so, most conversations linger somewhere between fear and enthusiasm where general AI is concerned – and that’s even before we consider a future potentially defined by autonomous, empowered AI agents. This was the sentiment at Davos this week, where AI took centre stage in discussions. In a blogpost accompanying the IMF research, MD of the International Monetary Fund (IMF) Kristalina Georgieva’s tone was tinged with foreboding and positivity. According to the IMF’s research, 60% of jobs in advanced economies such as the US and UK are exposed to AI and half of these jobs may be negatively affected; AI will affect 40% of jobs around the world; but AI could also “jumpstart productivity, boost global growth, and raise incomes around the world.”

More big questions. What’s the precise meaning of “affect”? And in this context, what exactly does “exposed” entail? The answers are still largely unclear, and there are both encouragingly creative and terrifyingly dystopian definitions of both. 

But Georgieva made one point that many would have guessed at, but maybe not wanted to acknowledge. In the past, routine tasks were the usual target for automation and technology, but what sets AI apart is its potential impact on highly-skilled positions; ordering a new pair of sneakers, for example, may be where AI agents start, but it won’t be where their deployment ends. Consequently, the major repercussions are anticipated to be felt not in the developing world, but in the developed one. Nevertheless, European Commission President Ursula von der Leyen was effusive when speaking on the importance of AI, stating that Europe’s future competitiveness “depends on AI adoption in our daily business.”

For now, it’s a waiting game for brands and retailers to see which of the approaches to AI proves to be the most powerful and popular, and brands and retailers can prepare by being open to adapt, adjust, and advance as things happen. 

The best from The Interline: 

Kicking off the week, we talk to SEDDI’s CEO on cultivating a culture of innovation, and why fashion must see digital product creation and digital product development as long-term commitments, not just short-term fixes.

The Interline & Harry Tribe explore the creative process, the software ecosystem, and the educational background that combined to create the visual identity for The Interline’s latest DPC Report – and ask what it all means for the future of digital product creation.

Neha Singh of OBSESS, discusses the emergence of a new, 3D-native retail channel presenting new possibilities.

In partnership with CGS, we chart historic British business Jain Group’s experience using B2B software to achieve further growth.