Pinterest Aims to Improve Discovery with Its AI Search Tools

Pinterest has added more AI options for visual search, hoping to drive discovery. The idea is to make it easier to segment elements in Pin image searches, building databases of relevant or personally associated products. When users view a Pin, they’ll now also see AI-generated words that can be used to learn more about elements of an image that interest them, including informationally and to facilitate shopping. “We’re breaking down and decoding images so users can quickly search and shop for the details of an outfit — whether it’s an aesthetic, color palette, fit or product category,” according to Pinterest.

The idea is to facilitate search returns based on specific visual elements of a Pin image. Pinterest also wants to make it easier to visually search individual image elements, expanding functionality of the home feed “Search Image” button by “enabling users to long-press on any Pin to switch to the visual discovery process,” reports Social Media Today.

Pinterest has also improved search customization, adding a “refinement bar” to help find items in a different color or fabric. “Use the ‘style’ refinement to see similar outfits but more y2k, or you love this blue blazer, and use the ‘occasion’ refinement to see more formal options,” Pinterest writes in a newsroom post.

“The company believes the new additions will help Pinterest stand out as a search engine for things users may otherwise have trouble describing with words — like a vibe, style, or aesthetic,” TechCrunch writes of the new implementation, initially rolling out in the U.S., Canada and the UK, focused on women’s fashion.

“A growing number of Pins on Pinterest will have specific pieces of an outfit or aesthetic highlighted with a shimmering animated glow,” notes TechRadar, explaining that if you “click on, say, the plaid skirt or the blue jacket, and Pinterest will show you some of the words to describe the item and help you find more like it without you needing a fashion dictionary.”

To implement the new features, Pinterest is tapping not only generative AI but also Visual Language Models (VLMs), TechRadar says. “The VLMs essentially translate images into searchable terms, decoding your taste from an image and helping you shop for it. It’s like a more nuanced reverse image search tool, one that can converse with you about your interest in vivid tones and stereopunk plaid.”

Generative AI hasn’t always been a plus on Pinterest, with users complaining about the proliferation of “AI slop.” The platform is launching a filter that will let users “select a ‘see fewer’ option on GenAI Pins in certain categories — like beauty and art — that are prone to being overrun with AI content,” TechCrunch reports, adding that “over time, it will expand this option to other categories, as well.”

No Comments Yet

You can be the first to comment!

Leave a comment

You must be logged in to post a comment.