Amazon’s new AI guides know what you want to buy before you do

October 11

Amazon is infusing AI into a new facet of its online shopping experience with the release of its AI Shopping Guides. The new feature aims to take over the product research part of hunting for a product by employing generative AI models that make personalized companions for your search. You can try it on Amazon’s website and mobile app, though just in the U.S. for now, and see guides for more than 100 categories of products.

The AI Shopping Guides start with a standard search for an item. Along with the usual list, you’ll see an AI guide appear if there is one for the category you’re looking at. When you open the guidebook, you’ll see an AI-curated collection of information about the products, including basic information, charts comparing different specs, and reviews from previous purchasers.

The AI models create guides that analyze all of the information Amazon has about the products available for sale and how people review them to write their own descriptions. It will even rewrite the descriptions so that you can understand some of the terms you might not know. The AI will then write custom recommendations for you based on its analysis of both the products and your own search and purchase history. Notably, using AI to keep scanning product listings means the AI Shopping Guides will stay up-to-date even as new products come out and old ones are updated or removed.

data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///ywAAAAAAQABAAACAUwAOw==

(Image credit: Amazon)

Rufus Writes In

The AI Shopping Guides also act to augment the Rufus AI chatbot, which Amazon deployed to offer AI-powered research and recommendations. Rufus makes the AI Shopping Guides interactive by answering follow-up questions you might have, personalizing your look for a product even further by, for instance, answering a question about two similar products by not only comparing them to each other but to products in a similar category you might have previously bought.

Rufus also includes some suggestions based less on your history and more on what companies have paid to have advertisements relayed by the AI. Amazon has also hinted at plans to further personalize its recommendations by using (with your permission) data from Gmail and YouTube.