\n
<\/p>\n
Walking down the street to scavenge for creatures like Snorlax and Squirtle was a regular activity in 2016, as the world was introduced to augmented reality (AR) games, thanks to Pok\u00e9mon Go<\/em>.<\/p>\n But now, 10 years later, images captured by players looking to “catch them all” are helping Niantic Spatial\u2014a spinout of Niantic, the San Francisco\u2013based company behind the game\u2014teach robots to navigate the world more effectively by building a highly accurate visual positioning system.<\/p>\n “It turns out that getting Pikachu to realistically run around and getting Coco\u2019s robot to safely and accurately move through the world is actually the same problem,\u201d John Hanke, CEO of Niantic Spatial, told<\/a> MIT Technology Review<\/em>.<\/p>\n According to MIT Technology Review<\/em>\u2019s recent report, Niantic Spatial partnered with Coco Robotics, whose flight\u2011case\u2011size delivery robots can often be spotted on the sidewalks of Los Angeles, Chicago, Jersey City, Miami, and Helsinki. But the success of the robots depends on them reaching their destination on time\u2014meaning not getting lost.<\/p>\n That’s where Pok\u00e9mon comes in.<\/p>\n When searching for highly coveted Pok\u00e9mon, players flocked to hot spots, taking multiple scans of the surrounding urban environments and capturing many angles of a single location.<\/p>\n These scans\u2014as well as those captured by players of Ingress<\/em>, another Niantic phone\u2011based AR game\u2014became some of the 30 billion images used to train Niantic’s current visual positioning model.<\/p>\n \u201cWe had a million\u2011plus locations around the world where we can locate you precisely,\u201d Brian McClendon, CTO at Niantic Spatial, told MIT Technology Review<\/em>.<\/p>\n Gathering enough data to create one of the world’s largest image data sets required an enormous manpower\u2014relying on players of a popular game to help. Just in its first week, Pok\u00e9mon Go<\/em> was downloaded by over 500 million people, still drawing in millions of users almost a decade later.<\/p>\n However, not all data is used for the mapping efforts, with players having to opt in to scan their surroundings. For instance, the data used to train the model received scans from the 2020 feature called “Field Search,” which gave players in-game rewards in exchange for scanning landmarks.<\/p>\n While many players may have opted in without understanding the implications, Niantic’s terms of service<\/a> is transparent about the company using user content to produce new services.<\/p>\n “Players have to choose to scan a specific public place\u2014it\u2019s not passive, and regular gameplay doesn\u2019t train any AI,” a Niantic spokesperson told Fast Company<\/em> when reached for comment. “We\u2019ve been transparent about this since 2019 in our privacy policy and public announcements.”<\/p>\n But while helping robots bring warm pizza to college students seems like a noble pursuit, not everyone is happy\u2014or surprised.<\/p>\n “500 million people played Pok\u00e9mon Go; scanned every street, building, and corner on earth; thought they were catching Pikachu. Niantic was building a 30 billion image AI map of the world now powering delivery robots that don\u2019t need GPS,” a user shared to X<\/a>. “You were the product the whole time.”<\/p>\n At the same time, many users claim to have seen it coming.<\/p>\n “Anyone who thought the AR task was there for anything other than collecting real-world data is oblivious to the real world. I fully expected it,” one user shared<\/a> on Reddit.<\/p>\n Another added<\/a>: “No shit, did you think we were scanning pokestops for shits and giggles?”<\/p>\n And others are taking the news with humor.<\/p>\n “Have you seen some of these poor bots trying to navigate,” one Redditor says<\/a>. “We didn’t do a good job.”<\/p>\n“We’ve been transparent about this”<\/h2>\n