How Project Genie Uses Street View to Simulate Real-World Places

Google DeepMind is expanding Project Genie, an experimental AI prototype capable of generating interactive virtual environments.

With its latest update, Genie can now connect its generative world-building capabilities with real-world imagery from Google Street View. This means users can create imaginative, explorable worlds anchored in real places.

For businesses, this update points to an important shift. AI is moving beyond text and images into interactive, location-aware environments that can help people explore, test, train and understand real-world spaces in new ways.

Bringing Real Places Into Interactive AI Worlds

Project Genie is designed as a general-purpose world model. It can generate diverse, interactive environments that AI agents can move through and learn from. Google says Genie has already been used for research, including helping agents learn and reason in complex virtual settings.

The latest update takes this further by connecting Genie with Street View.

Users can choose a real place in the US, select a visual style and describe a character. Genie then creates an imaginative world with a starting point grounded in Street View imagery.

For example, a user could explore a real-world landmark with an underwater theme or reimagine a historic location in a different visual style.

The result is an interactive space that starts from a real-world location.

Why Street View Grounding Matters

Street View grounding gives AI-generated environments a stronger connection to reality.

Instead of creating a world from a prompt alone, Genie can use real-world imagery as a reference point. This makes the generated environment more context-aware and more connected to an actual place.

For businesses, this is where the idea becomes especially interesting.

Real-world imagery, maps and location data can help teams better understand physical environments before they visit, develop, service, deliver to or operate within them.

This kind of technology points to future use cases in which organisations could simulate real-world environments, explore operational scenarios, and make better location-based decisions.

What This Could Mean for Business Applications

Project Genie is still experimental, but the direction is relevant to many industries.

• For logistics and transport, simulated environments could support better route planning, driver training or scenario testing.

• For property, infrastructure and field services, visual location context could help teams understand a site before sending people into the field.

• For retail and customer experience teams, interactive location-based tools could help customers explore spaces, branches or service areas in a more useful way.

• For ISPs and infrastructure providers, the broader lesson is that location data becomes more valuable when it is easy to visualise, explore and act on.

The key is not only having the data. It is making that data understandable.

Maps Imagery Grounding and Developer Potential

Google says the new Street View capability is powered by Maps Imagery Grounding, the same technology developers can use to create AI visuals with Street View.

This is an important development for organisations exploring how Google Maps Platform can support richer digital experiences.

When developers can combine imagery, location intelligence and AI, they can create tools that go beyond static maps. These tools can help users understand places more clearly, whether they are:

• Planning routes

• Assessing sites

• Reviewing service areas

• Improving customer-facing applications

For 28East customers, this reinforces the value of building on strong location foundations.

Good maps, accurate imagery and well-implemented APIs can support better experiences across sales, operations, logistics, support and planning.

What This Means for South African and African Businesses

Across South Africa and broader Africa, businesses often operate in environments where location context matters.

Between complex routing, unpredictable addresses, and shifting infrastructure, managing operations on the ground is a massive juggling act. Yet, despite all this backend complexity, customers still expect a flawless, transparent digital experience before they're willing to commit to a purchase.

AI-generated environments are still emerging, but the underlying principle is already practical.

The better a business can represent the real world digitally, the easier it becomes to plan, sell, support and operate.

This starts with reliable location data, strong Google Maps Platform implementation and practical tools that help people make better decisions.

28East Helps You Build Better Location-Based Experiences

At 28East, we help businesses use Google Maps Platform, APIs and location intelligence to build digital tools that are useful and aligned with real operational needs.

Whether you need interactive coverage maps, address verification, route optimisation, Google Maps support or help understanding which APIs fit your use case, we can help.

Our team combines technical Google Maps expertise with local support and a practical understanding of South African and African markets.

Let’s Make the Real World Easier to Work With Digitally

Project Genie shows how AI and real-world imagery can make digital environments more interactive and more useful.

For businesses, the opportunity is clear. Better location intelligence can help teams understand places, improve decisions and create digital experiences that feel more connected to the real world.

If your business relies on maps, routes, addresses, service areas or site context, 28East can help you turn location data into tools that support better outcomes.

(Source: https://blog.google/innovation-and-ai/models-and-research/google-deepmind/project-genie-expands/)