iStockphoto / Wachiwit
It has been around 10 years since Pokémon Go peaked in popularity. But there was a time when millions of people worldwide were pounding the pavement in search of pocket monsters. Heck, there was a stampede in Taiwan that brought traffic to a standstill from people trying to catch a Snorlax.
With 10 years of real-world data amassed, with users taking billions of photographs of their surroundings, walking real-world routes around homes, office buildings, etc., it has now become evident that the data amassed was gathered to train AI systems.
How Pokémon Go Data Is Powering AI Delivery Robots
After Niantic amassed over 30 billion real-world images, the largest real-world robotics perception dataset ever created, they have since turned to using that data to enable delivery robots to navigate the real world. Looking back, the ‘Rise of Artificial Intelligence’ was barely a blip on the radar a decade ago. Few people were taking it as seriously as the vast majority of the tech industry is today.
Niantic, the maker of Pokémon Go, must have been thinking ahead in ways nobody anticipated because they sit on the largest treasure of real-world mapping ever created.
Going back to November 15, 2024, Niantic made their intentions clear. In a blog post titled ‘Building a Large Geospatial Model to Achieve Spatial Intelligence’ they announced how the data gathered within the app, from users taking pictures of their pokémon and navigating the world around them, would be used to power robots navigating through the real world.
In that blog post, Niantic announced:
“As part of Niantic’s Visual Positioning System (VPS), we have trained more than 50 million neural networks, with more than 150 trillion parameters, enabling operation in over a million locations. In our vision for a Large Geospatial Model (LGM), each of these local networks would contribute to a global large model, implementing a shared understanding of geographic locations, and comprehending places yet to be fully scanned.”
Flash forward to now and this video from Newsforce explaining how all of the data was amassed and is being used has nearly 17 million views in 24 hours. Everything said in the clip is outlined above but it is still worth watching:
POKÉMON GO PLAYERS TRAINED 30 BILLION IMAGE AI MAP
Niantic says photos and scans collected through Pokémon Go and its AR apps have produced a massive dataset of more than 30 billion real-world images.
The company is now using that data to power visual navigation for delivery… pic.twitter.com/FIs65uO3sx
— NewsForce (@Newsforce) March 15, 2026
As of March 2026, there are still an estimated 5.4 to 5.7 million daily users of Pokémon Go. The game has been downloaded over a billion times and at its peak there were an estimated 45 million daily users.
All of those people were walking around through the real world… Navigating sidewalks, trails, hallways, stairwells, department stores, office buildings, schools, etc. They were taking photos of the pokémon they capture as that was often one of the challenges in the game required to unlock additional levels.
In the wrong hands, that data would be used in dangerous ways. But when it comes to AI-powered delivery robots needing to navigate passages unfamiliar to traditional robotics datasets, they are sitting on the largest gold mine of data in the world.
Here is how Niantic imagined that data being used when they announcedthis project back in November 2024:
“As humans, we have “spatial understanding” that means we can fill in these details based on countless similar scenes we’ve encountered before. But for machines, this task is extraordinarily difficult. Even the most advanced AI models today struggle to visualize and infer missing parts of a scene, or to imagine a place from a new angle. This is about to change: Spatial intelligence is the next frontier of AI models.”
Is this the future? Let us know what you think in the Facebook comments.