What should I wear today and then for a formal cocktail party after work? Where can I buy it?

How much will it cost? These queries typically stump even the best retail search engines by returning half answers or error messages. 

Mastercard’s Dynamic Yield wants to change that. It developed the generative artificial intelligence (GAI) tool, Shopping Muse, fed with first-party data from the retailer’s website.

“Shopping Muse uses a combination of contextual insights – like location, when enabled, device type like mobile and desktop, and real-time behavior such as what items the shopper is clicking on across the site and the questions they’re directly asking the Muse – to personalize its recommendations,” wrote Ori Bauer, CEO of Dynamic Yield by Mastercard, in an email to Search & Performance Insider.

He said it also can leverage shoppers’ demonstrated affinity from past purchases or saved items, for example, to tailor the experience based on each shopper’s preference for particular brands, sizes, colors, and more.

The tool recreates an in-store experience using a human personal shopper by translating consumers’ casual search terms, such as “Find me an outfit for New Year’s Eve,” into product recommendations. It goes beyond the basics to suggest coordinating products and accessories.

Unlike other conversational search tools, Shopping Muse recommendations are personalized to match an individual consumer’s unique profile and intent, and build on the context of conversations over time. The results are designed to “perfectly match even the most eccentric query.”

Mastercard said the tool goes beyond keywords — translating visual attributes of a product into data so it can recommend options even if they’re not tagged by a keyword in the product feed.

In addition to helping shoppers search by phrase, Shopping Muse can help consumers find the perfect item even when they don’t know how to properly describe it in words.

Using integrated advanced image recognition tools, retailers can recommend relevant products based on visual similarities to others, even if they lack the right technical tags.

On top of these proprietary algorithms, the solution leverages third-party Large Language Models (LLMs) and image recognition algorithms to perform various tasks. Shopping Muse does not leverage Mastercard data, Bauer explains.

The tool takes into account the shopper’s likes and dislikes based on session browsing history or past purchases, to better estimate future buying intent. With an understanding of the consumer’s affinity and the context of broader collective behavior, the retailer can ensure the suggested items are complementary, not redundant.

Although Mastercard developed Shopping Muse for the fashion industry, other industries and categories could use the technology.

A furniture store could use it to help consumers decorate a room by querying the dimensions of the space and adding the specific items required. For example, “would this blue colored furniture fit in a 12×12 room?” 

A grocery store could use Shopping Muse to help plan a holiday dinner party or week-long menus for a family of four. 

GAI will help brands, if they are not using it already, create new products and services, determine color and weight. Determine the best media to buy after creating an ad with copy that will resonate best with a certain type of consumer.

Some of us, me, are waiting for the super intelligent tool such as the agents that Bill Gates described in a blog written earlier this month, but Microsoft President Brad Smith told Reuters there is “no chance” a super-intelligent AI tool, where computers are more powerful than people, will be crated in the next 12 months. It could even take “years” or “decades” before that becomes a reality. Safety and controls need to be put in place.

Nvidia CEO Jensen Huang, speaking at the New York Times DealBook Summit (December 26, 2023), said AI is “gaining on humans” and will likely be “fairly competitive” with humans in five years.

In an interview, OpenAI CEO Sam Altman spoke briefly to The Verge about an “unfortunate” leak pertaining to Q*. Some suggest that Altman’s firing at OpenAI may have been due to a lack of transparency about a secret project known as Q*.

“No particular comment on that unfortunate leak,” Altman told The Verge. “But what we have been saying — two weeks ago, what we are saying today, what we’ve been saying a year ago, what we were saying earlier on — is that we expect progress in this technology to continue to be rapid and also that we expect to continue to work very hard to figure out how to make it safe and beneficial.”

Yejin Choi, a computer science professor at the University of Washington, has her own unique view about how AI gains knowledge and what to expect in the future.