Google Maps adding generative AI to 'supercharge' search

Generative AI is coming to Google Maps and will let you find places in a conversational manner.

Google is tasking its large language models (LLMs) to “analyze Maps’ detailed information about more than 250 million places,” as well as user-generated photos, ratings, and reviews from 300 million contributors, to find locations that match your prompt.

This is currently an experiment, but Google today shared a straightforward “Ask about” UI.

In the example below, searching for “places with a vintage vibe in SF” returns vinyl record stores, clothing stores, flea markets, and other high-level categories. You get basic location details and review summaries that “highlight why a place might be interesting for you to visit.”

You’re able to “Ask a follow up,” like “how about lunch,” that retains the context of the original query: “Maps will suggest places that match the vintage vibe you’re looking for, like an old-school diner nearby.”

For example, if you’re out and about and it starts to rain, just ask Maps for “activities for a rainy day” to get suggestions for indoor activities. You’ll get helpful suggestions like comedy shows or movie theaters. If you’re looking for options the whole family can enjoy, ask a follow-up question like, “What about options for kids?”. You’ll see places like bowling alleys, children’s museums or indoor playgrounds depending on where you are, along with highlights based on reviews from people who have been there.

Other possible prompts include:

Best brunch spots nearby
Things to do for outdoor enthusiasts near me
Fun things to do with friends nearby

This experimental generative AI capability is rolling out this week to US users in the Google Maps Local Guides contributor program. It will be available for “everyone over time.” Google says this is the “beginning of how [it’s] supercharging Maps with generative AI.”