5 Years After Pokémon Go, It's Time for the Metaverse

Plus: The biggest idea of 2016, the ethics of synthetic data, and a blaze on the Gulf of Mexico.
People in Pokemon hats looking at phone
Pokémon GO was the first wildly popular implementation of augmented reality, giving us a preview of what techno-pundits now believe is the next big thing.Photograph: Tomohiro Ohsumi/Getty Images

If you buy something using links in our stories, we may earn a commission. Learn more.

Hey, everyone. What if they had an Olympics and nobody came? I guess we’ll find out.

The Plain View

The summer of 2016 was only five years ago, but it feels like a century has passed since then. Despite a radically unconventional presidential candidate (yes, I am restraining myself), we were luxuriating in what we now longingly refer to as normalcy. With one exception. Millions of people were taking to the streets and parklands holding their iPhones to their faces so they could see what wasn’t really there—a zoo’s worth of cartoony creatures. Through an app, users not only saw those creatures on sidewalks, park benches, and stadiums, but they could take steps to “capture” them and then pit those beasties against other critters in competition.

That’s right, it’s now the fifth anniversary of Pokémon Go and the craze that marked its launch. That phenomenon was not only a milestone for the company behind the game, a Google offshoot called Niantic, but for the digital world in general. Pokémon Go was the first wildly popular implementation of augmented reality, a budding technology at the time, and it gave us a preview of what techno-pundits now believe is the next big thing.

To mark the anniversary, I had a conversation with Niantic CEO John Hanke, who founded the company after heading Google Maps. Appropriately, he conducted our Zoom chat as he sauntered through the streetscape of historic downtown Truckee, California, staring into the phone like I was a Pidgeot or a Tangela that he was about to capture.

“We benefited from the huge social meme-ification,” he says of that breakout 2016 summer. “And then there was the backlash meme-ificiation of Pokémon Go/Pokémon Gone.” Indeed, the mania subsided, as most players got their fill. But a solid core of fans have kept the game alive (422B Pokémon caught to date!), and Hanke says that Niantic is profitable. This month, the company will host its annual Pokémon Go Fest, with rare characters in the wild, abundant bonus points, and new electro-pop soundtrack music; last year millions attended virtually. I confess that I am one of those Pokémon Goners who haven’t logged on in years, but awakening the app again I found tons of new features. Did you know you could release a Poké critter and take a selfie with it?

Pokémon Go is just the first step in Hanke’s larger vision of welding digital artifacts to the real world. While Niantic has concentrated on gaming and entertainment to date—its current and upcoming products are built on the respective worlds of Harry Potter and Transformers—Hanke believes he is constructing what he calls the Real-World Metaverse. Science fiction fans will recognize the term “metaverse” as the shared geographic hallucination that novelist Neal Stephenson introduced in his book Snow Crash—an alternate world that people could dive into via virtual reality. But unlike the total, Matrix-like immersion into an imaginary realm that virtual reality offers, Niantic’s Metaverse lives on top of the actual world, enhancing specific locations with digital roadside attractions.

Niantic isn’t alone in this enterprise. It seems that every big tech company, particularly those worth a trillion dollars or more, has become infatuated with the possibilities of augmented reality. Right now, the main way people access AR is by holding phones up to their faces, but the nerd cognoscenti all agree that one day special eyeglasses will persistently offer an augmented universe. Snap and Microsoft already have products on the market, Facebook is openly developing its own AR spectacles, and Apple is rumored to be working on glasses, too. Even the deflated fortunes of one vaunted AR company, Magic Leap, haven’t slowed the excitement. (Niantic is among several companies that have scooped up some of its employees.)

It would seem that Niantic is at a disadvantage in competing with those huge players. Besides more limited resources, it doesn’t have the leverage of a dominant operating system or a giant social network. But it does have a plan. Niantic is allowing developers to use its technology to create their own AR applications, some of which we may see later this year. This means sharing the company’s crown jewel, known as the Map—a hyper-accurate positioning system that ties digital objects to real-world geography. The company has also partnered with Qualcomm to create a reference standard for AR glasses, one available to any company that wants to manufacture a lightweight augmented reality rig. “I don’t think you need to be a giant to lead innovation,” says Hanke.

Creating the Real World Metaverse is like expanding the lot of every building in the world to accommodate unlimited additions, without a building code to constrain the possibilities. But it opens up a lot of questions, too. Let’s assume AR apps become common, and we routinely see and hear virtual objects as we navigate the world. Will we be locked into Apple’s or Facebook’s metaverse depending on whose device is augmenting our senses? There are also legal issues. Who owns the virtual airspace of a physical location? The success of Pokémon Go has already thrust Niantic into the thick of these seemingly far-fetched debates. It settled a case from aggrieved property owners who unwittingly hosted virtual Pokéstops, by promising to avoid augmentation of single-family homes and responding quickly to complaints.

Is this sounding like science fiction? It’s no coincidence. Hanke told me Vernor Vinge’s 2006 novel Rainbows End “has been kind of my North Star for AR.” In that book, augmented reality is ubiquitous, delivered through contact lenses and smart clothing. Vinge also has an explanation of how alternate versions of geographically linked AR worlds can coexist: He envisions “belief circles” where people choose which set of enhancements augment the world they move around in. Uh-oh. It’s become a cliché to say that some people deny the truth right in front of their eyes. But what if their eyes were fed the fantasies they chose to believe?

Yes, 2016 sounds ancient. But we may remember it as the first time a mass population got a taste of one of the metaverses that would become our natural habitat. Rainbow’s End, by the way, is set in 2025.

Time Travel

In a December 2016 Backchannel story, I shouted out Pokémon Go as the year’s Big Idea:

Introduced on July 6 by a then-obscure company called Niantic, Pokémon Go *saw the most insane, huge adoption of a mobile game—ever. Well before summer’s end, 500 million people had downloaded it, rocketing the free app to the top of Apple and Android stores and suddenly turning it into the center of people’s daily lives. Though “augmented reality” was still an ominous buzzword in the tech world—with Microsoft showing an early prototype, Facebook and Apple rumored to be working on it, and a mysterious company in Florida struggling to set a launch date for a system it says will rule them all—*Pokémon Go made it happen now, inviting the popular 1990s dramatis personae of the Japanese card game to share our world with us and daring us to capture them by tossing virtual beach balls at them … 

But it isn’t the number of people who downloaded the app and made themselves look like idiots while flipping Pokéballs at imaginary creatures that led Backchannel to choose Pokémon Go—along with Niantic and its founder and CEO John Hanke—as the recipient of our first annual Big Idea award. Though clearly the gaming phenomenon of the year, Pokémon Go is more than a fad, more than a game, and more than an app. It is a harbinger of a future where reality is malleable, where the physical world has more annotations than Infinite Jest, and where computer-coded apparitions intrude in the firmament to entertain, inform, and alter our behavior.

Ask Me One Thing

Jenn writes, “I’m hearing more people rally around synthetic data for AI training. Do you think this will solve bias problems or create new issues we haven’t anticipated?”

Excellent question, Jenn. First, a quick explanation of synthetic data. When teaching a computer to master human tasks like understanding speech or identifying cats and dogs in images, the preferred technique is to “train” the system using big data sets—a huge corpus of information drawn from the real world. You might feed all the transactions from a credit card company, or the photo libraries of millions of users, into the system so it could learn how to identify, respectively, frauds or faces. One downside of using actual examples is possible privacy violations or, as you imply, potential biases passed on from the training examples. Synthetic data uses simulated examples, generated by AI, to train the systems. Since the data is all computer-generated, this solves privacy issues, and it’s often cheaper. But it doesn’t necessarily eliminate bias. Those artificially generated examples have to begin with some premises, and it’s possible that those might have inadvertent bias.

No matter how we train our AI, we have to be vigilant in identifying bias. No shortcuts.

You can submit questions to mail@WIRED.com. Write ASK LEVY in the subject line.

End Times Chronicle

Hey, remember when the Gulf of Mexico caught fire? Last week?

Last but Not Least

A report from the front lines of trans medical care—prison.

And while we’re visiting front lines, here’s a dispatch from the Kansas lab charged with preventing bio-tech attacks from agriculture-based pathogens.

Netflix is losing its mojo. Maybe hire Ted Lasso as a coach?

I have seen the future of dance and its name is … Boston Dynamics?

More Great WIRED Stories