Google’s Revolutionary Project Genie: Transforming Street View into an Interactive Experience
We’ve all used Street View on Google Maps to show a friend what our childhood home looked like, or dropped that little icon of a person onto the streets of Paris to see if we’d booked a hotel in a cool neighborhood. Imagine being able to do that, but in a more immersive and interactive way that allows you to truly simulate the street and its surroundings, and even do things like adjust the weather or see what it would look like in a “Day After” scenario.
Introducing Project Genie at Google I/O 2026
This is one of the goals of Google’s latest integration. Starting today, Google DeepMind connects Street View to Project Genie, the company’s general-purpose global model that can generate diverse and interactive environments. The new feature launched at the Google I/O 2026 developer conference.
“It’s really powerful for both the agent [and robotics] use cases and for humans to play with, and that’s always been Genie’s thesis,” Jack Parker-Holder, a research scientist on DeepMind’s open team, told TechCrunch.
Transforming Robotics and Human Interaction
Jack Parker-Holder provided an example involving a new robot deployed in London, which rarely sees the sun. According to Parker-Holder, Genie could simulate those rare occasions when the sun shines on Victorian houses, so that the rays don’t shock the robot when that happens.
“At the same time, you might say, ‘I’m going to New York, but not at this time of year,'” he continued. “‘It’s going to snow. I want to see what this block looks like in the snow.'”
Building on Two Decades of Street View Data
Google has been collecting Street View data for 20 years via cars equipped with cameras and individuals attached to “tracking backpacks.” The tech giant has collected more than 280 billion images from 110 countries and seven continents.
“With Street View, we have images from a large part of the world,” Jack said. “You can imagine how potentially powerful it is to combine this rich source of real-world information and data with the ability to simulate worlds.”
Genie 3: A New Era of Interactive Environments
Google released its latest Genie 3 global model for a search preview last August and opened access to the tool to Google AI Ultra subscribers in the United States in January, allowing customers to create interactive game worlds from text or image prompts. The goal is to use Genie for educational experiences, games, and robotics training.
Genie 3 is already helping power one of Waymo’s simulators to train its self-driving cars on “extremely rare events” like tornadoes or occasional encounters with elephants. Adding Street View data could help Waymo prepare for its launch in more cities around the world.
Waymo has its own simulator that it has relied on to scale to 11 U.S. cities and test its AI pilot in several others. The difference with Genie, Parker-Holder explains, is that it’s all from the car’s point of view. Street View not only allows you to simulate a world anchored in a real place, but also to move the point of view towards other types of agents, such as a human or a robot.
Expanding Access and Future Prospects
Google is launching Street View in Genie today for select Ultra users in the US, with access rolling out broadly over time. Global Ultra users will have access to it in the coming weeks, according to the company.
The researchers’ goal is to put this new capability in the hands of as many people as possible, according to Diego Rivas, product manager at DeepMind. He cautioned that Street View in particular and Genie in general are still an experiment and so there is much room for improvement in terms of accuracy.
In the samples the Google team showed me — including an underwater simulation of a neighborhood I lived in — the results are impressive and recognizable, but remain video game quality rather than photorealistic. The models are not yet sensitive to physics, which means they do not yet understand cause and effect. For example, in a simulation of a woman running through a snow-covered Joshua Tree, she ran through cacti and bushes.
Compare this, for example, to Google’s Nano Banana image generator – which can now generate perfect text in infographics – or its Veo video generator – which understands that paper boats drift in water currents, smoke disperses in the air and fabric covers shapes.
Physics is not hard-coded into these models; they learn it intuitively over time through passive observation, as a living being would.
“I think for this type of model, there’s maybe six to 12 months behind video in terms of accuracy and quality, so I think that’s something we’ll address,” Parker-Holder said.
Jonathan Herbert, director of Google Maps who started on the Street View team as an intern 12 years ago, said Genie could not yet create a faithful reconstruction of a street. He believes that the real breakthrough lies in the spatial continuity of AI. If you turn 360 degrees, the AI remembers and correctly simulates the environment behind you. From there, the model can create a new environment on top of it.
“We have been thinking for a long time about how we can create the best and richest model in the world from Street View data,” Herbert said. “We’ve had the idea of using map data in new ways and for new types of AI research for a long time.”
Keep up to date with the rest of the big news from Google IO 2026
Google Search as you know it is over
Google updates Gemini app to support ChatGPT and Claude
Google Introduces Gemini Spark, a 24/7 Agent Assistant with Gmail Integration
How to use Google’s new information agents
When you purchase through links in our articles, we may earn a small commission. This does not affect our editorial independence.
Source: Here
“`
