Google Maps has long been far more than just a navigation app. Its extensive database of locations, reviews, photos, and local insights has enabled people to discover hidden gems in cities around the world for over 15 years. Yet even with hundreds of millions of places mapped, there are still amazing spots waiting to be uncovered. To aid further discovery, Google is now augmenting Maps with an AI-powered recommendation engine.

Powered by large language models (LLMs) that can understand nuanced natural language prompts, this new AI integration aims to provide personalized and creative suggestions tailored to each user’s tastes, interests, and explore-worthy whims.

Say you find yourself with a free afternoon to thrift shop during an upcoming trip to San Francisco. Simply open up Maps, type “places with a vintage vibe in SF,” and let the on-device AI work its magic. After analyzing Maps’ riches of data — 250 million-plus business listings, imagery, ratings, reviews, attributes, and more — as well as your own search history and past preferences, you’ll be served up delightful vintage store options to fill an afternoon, each complete with photos, reviews and useful local gems only the AI assistant could piece together.

Far more than just rehashing what you could have found through a standard keyword search, the AI promises human-like recommendations, with convincing reasons why each shop is worth a visit. The technology is still in early days, as are all generative AI applications, but Google is focused on developing this responsibly.

For now, only members of Google’s Local Guides program will get to experience and provide feedback on the AI integration during its preview period. These dedicated explorers and content creators for Google Maps are ideally suited to ensure such novel technology builds upon the app’s strengths as a discovery engine without amplification issues.

As AI continues progressing at astounding rates, Google Maps stands ready to tap that potential for personalized search and recommendations — changing how we explore the world in endlessly serendipitous new ways. Today, it’s about finding hidden vintage shops in San Francisco. Tomorrow, who knows what unexpected delights we’ll discover? One thing’s for sure, though: the future of recommendations is AI-powered.