Right then. Here’s something nobody saw coming, or perhaps we just weren’t looking hard enough.
Millions of people spent years wandering pavements in blistering heat, pointing their phones at park benches, statues, and oddly specific building corners. We thought we were catching digital creatures with silly hats. Turns out, we were building the most sophisticated pedestrian-level mapping system on the planet—and getting paid in imaginary currency worth precisely nothing outside the app.
Rather brilliant, when you think about it. Though “brilliant” rather depends on which side of the transaction you’re standing.
The Mechanism
The mechanism was elegant in its simplicity. Niantic introduced an “optional” scanning feature—voluntary, you understand, in the same way that breathing is voluntary if you don’t mind the consequences.
Players were offered:
- Pokéballs
- Raid passes
- The occasional Ultimate Potion
What they received in exchange for their labour was digital trinketry digital trinketry worth nothing. What Niantic received was something altogether more valuable: over 30 billion street-level photographs, captured from every conceivable angle, in every weather condition, across virtually every urban environment on Earth.
The scale is difficult to comprehend. Google Street View relies on fleets of vehicles with mounted cameras. Niantic discovered something far cheaper: human beings who would work for free if you gamified the experience sufficiently.
- No petrol costs
- No fleet maintenance
- No payroll
Just millions of enthusiastic volunteers mapping every alleyway and side street that a car could never reach.
From this mountain of visual data emerged what Niantic calls a “Large Geospatial Model”—think of it as ChatGPT’s cartographically obsessed cousin. Where language models predict the next word in a sentence, this model predicts physical space. It understands environments. It navigates. It knows that a doorway is a doorway regardless of lighting conditions.
The Endgame
The endgame became clear in 2025, when Niantic Spatial spun off as an independent entity and partnered with Coco Robotics. You’ve likely seen their little delivery robots—the suitcase-sized contraptions trundling along pavements, ferrying takeaways to hungry customers.
What you probably didn’t know is that these machines were struggling with something rather fundamental: GPS is notoriously unreliable in urban canyons.
- Tall buildings create signal shadows
- Satellite positioning drifts by metres when surrounded by glass and steel
- A few metres of inaccuracy means the difference between successful delivery and collision
The solution? Those billions of photographs you snapped whilst hunting Pokémon. The robots now use their onboard cameras to compare real-time visual input against Niantic’s database. Your Sunday afternoon stroll to catch a virtual monster taught a robot how to find its way home.
The Pattern
There’s a pattern here worth noting. Companies have discovered that the most valuable commodity in the artificial intelligence age isn’t code, or processing power, or even expertise. It’s data—specifically, data that humans generate through their everyday activities.
The question isn’t whether this is clever business practice. Demonstrably, it is. The question is whether it’s ethical to obscure the true purpose behind entertainment.
Niantic’s terms of service presumably covered this possibility in paragraph forty-seven, subsection C, buried beneath enough legal verbiage to bore anyone into clicking “accept” without reading. Whether buried disclosures constitute informed consent is rather another matter.
Where We Stand
So here we are. The future of autonomous delivery navigation was built by people chasing digital creatures through parks. They were never told the full picture. They were compensated in pixels. And the resulting technology will likely generate billions in value for shareholders who never took a single photograph.
One wonders whether the next great AI dataset is being built right now, hidden inside some innocuous app that millions of people use daily without a second thought:
- The gym check-in you just logged
- The route you mapped on your running app
- The photo you uploaded to that editing platform
Each leaves a data trail. Each has potential value beyond what’s immediately obvious.
Questions Worth Sitting With
If you’d known the true purpose of those scanning tasks, would you have still participated? And if so, what would have felt like fair compensation for your contribution to a multi-billion-dollar navigation system?
The gap between what companies know about user behaviour and what users know about company intentions seems to widen each year. Is informed consent even possible when the ultimate use of data hasn’t been invented yet?
What other everyday digital activities are building value that we’ll only recognise in retrospect—and who should share in that value when it materialises?
Leave a Comment