Pokemon GO data is used to train the geospatial AI model

Key insights

Niantic is building a geospatial AI model using data from popular AR games like Pokemon GO.
Called the Large Geospatial Model, the model aims to improve interactions in physical spaces worldwide.
The technology has a wide range of applications in AR glasses, robotics, content creation and autonomous systems.

Pokémon GO Developer Niantic has announced that it will use data from its popular AR games to create a new geospatial AI model that will enable machines and neural network interfaces to better interact with physical spaces around the world. Pokémon GOan AR spin-off of the popular franchise, uses players' smartphone cameras in conjunction with neural networks to place virtual objects in real spaces.

Niantic founder John Hanke led Google's geospatial division, which developed Google Maps, Earth and Street View, before founding Niantic in 2010. Previously, his company Keyhole created an interactive interface for live mapping conflicts in the Iraq War. Niantic is best known for its series of popular augmented reality games, including Pokémon GO, Ingress Prime, Pikmin Bloomand more.

Niantic announces “Next 10 Years” of Pokemon GO

A senior Niantic official teases “the next ten years” of Pokemon GO and reveals a surprising fact about the game.

As stated in a company blog post, Niantic's latest endeavor is the construction of a “Large Geospatial Model” (LGM), which combines the more than 50 million neural networks the company has trained into a single model of geographic locations and their spatial configuration would combine. The company cites “AR glasses…, robotics, content creation and autonomous systems” as key applications of this technology.

Niantic looks to the future with new technology investments

Niantic is banking on the rise of wearable XR technology and autonomous machines to make spatial computing the dominant operating system of the near future. The company's current geospatial models are based on millions of scans created by users of its AR games and Scaniverse app. These resources were combined to create its visual positioning system, which allows the placement of persistent virtual objects in real space with a high degree of accuracy. The Large Geospatial Model is the next evolution of this system, combining the millions of neural networks generated by the VPS into a single unit capable of extrapolating unscanned location information from local spatial and position data.

While Pokémon GO Users probably won't be surprised that the company behind a popular game is exploiting their data to pursue larger business goals. The implications of a large geospatial model are far-reaching.

As large geospatial models continue to evolve, it seems inevitable that Niantic or another company will explore the full range of possible integrations of the new technology. It remains to be seen what regulatory or ethical limits will be placed on implementing this remarkable advancement and what benefits and drawbacks consumers might experience with this latest AI development.