Google Maps is adding generative AI features to help users find destinations using conversational search in a limited experiment for users in the United States that is rolling out this week.

Detailed in a blog post published Thursday, the enhanced search in maps uses large language models that analyze map data from 250 million places and “trusted insights” from its community of over 300 million contributors to make suggestions on where to go.

Beginning in the U.S. in early access, select members of the Google Maps community dubbed “Local Guides,” who are among the most active and passionate users will gain access to the new system.

Using the power of AI, users will be able to open up a dialogue in maps and ask questions about the local scene such as, “Where can I go to find a vintage atmosphere?” The maps interface will then pass the question along to AI models to analyze information on local businesses, flea markets, thrift stores and the like.

Users will be able to see the locations found by the AI model on the map, select them and save them to a list for later. The results can be shared with friends so that groups can collaborate on outings or events.

The interface will even include images of the locations, helpful reviews from community members and summaries of how each location fits into the “vintage” question. The user can even follow up with extra queries such as, “How about someplace to eat?” Where the AI will include restaurants in the area that include an old-timey feel. Further follow-up questions could help reveal if eateries have vegan options and assist with filtering by cost and quality.

Google said that the follow-up feature will make it perfect for all the times when plans change unexpectedly due to a sudden rainstorm or a location being closed. Users can simply jump back to Maps and chat with the AI again to find something new as easily as asking a new question.

This is currently a limited access experiment available only to the local guides, Google said that it would be taking advantage of the guides’ expertise and feedback to help shape its development so that it could be rolled out to everyone over time.

Image: Google