Snapchat has made two incredibly smart business moves in the last year. The first was, obviously, enticing the Miami producer DJ Khaled to spend hours a day spouting his The Secret-like major keys (🔑) to success and getting lost on his Jet Ski. The second came last autumn, when the company acquired a Ukrainian start-up called Looksery to build out its wide slate of goofy, charming, and occasionally clunky lenses, which superimpose cartoonish effects onto users’ faces in real time. The most famous of these currently is the face swap, which (theoretically, under perfect conditions) will “swap” the faces of two people in a photograph or video.
The filters in general are extremely popular — not popular enough that people wanted to pay for them, but popular enough for Facebook to take notice and purchase a start-up doing brisk traffic in a face-swap app: MSQRD. (We wrote about MSQRD two months ago, when it was taking Russia by storm for a feature that allowed users to superimpose Stalin’s face onto theirs. It’s now among the most popular apps in the App Store.) Zuck announced the acquisition in an Iron Man face-swap video that he posted to his Facebook page:
The sudden rise of popularity in face-swap images is fascinating in part because they’ve always been a background part of the internet landscape, since at least the dawn of what we’ll call the “viral web” sometime in the 2000s. The premise is simple; as I say above, two people swap faces. Usually it’s funniest when the two people are dramatically different in age or appearance, like a baby and an adult. That’s funny. Thanks for letting me explain why something is funny. Like you, I enjoy comedy.
But if you wanted to make a competent face swap five or six years ago, you needed — at the very least — expensive photo-editing software and the skill to use it. The production of face swaps, until relatively recently, required some degree of financial means as well as expertise. Snapchat’s filter does away with all of that. Just get two faces in front of the lens and let the mysterious optical-recognition software work its magic. We now all have at all times, in our hands, the ability that was once the sole domain of the weird kid in class whose dad’s computer had Photoshop. More than anything else, this is behind the renaissance in face-swap images: Meme consumers are now meme producers.
The face swap isn’t the only instance of a once-niche meme made mainstream. Sites like Meme Generator and Quickmeme have long existed as factories and clearinghouses for producing instances of memes, as did the later Ragemaker.net. Web-native culture, in many ways, stems from these browser-based tools, creating homogeny. Memes democratize humor, but sites like these democratize their production.
More and more frequently, developers have realized that being the first to lower the technical barriers to meme creation can pay off in big ways. When Kanye West revealed the cover art for his latest album, The Life of Pablo, many quickly realized its meme potential — its template was simple and easily imitable. The unveiling also set off a very slight arms race to become the official unofficial TLOP cover generator. There was tlopcover.com, thepablo.life, and thelifeofpablol.com; all providing simple tools that West himself seemingly took advantage of. If you searched popular source-code repository Github, you could even find the resources to construct your own generator. We’re now firmly in a period where creating memes is not the only way to gain attention — the right tools to encourage their manufacture can also make or break a trend.
But there’s a reason that the face swap in particular has become an inescapable genre of viral image: Because it’s the perfect embodiment of our anxieties about technology growing in power and ability. When the face-swap filter works, it’s exciting and a little eerie. After all, the exact method through which Snapchat is able to transpose faces in real time is for all intents and purposes a black box of tech magic. Maybe it makes you, as it does me, slightly uncomfortable to think about how the device in your pocket is capable of automatically altering your appearance so completely and so realistically that you can be fooled, even for half a second, into thinking that your own face is in the right place on someone else’s body. The moment you realize something’s off — the moment you dive back down into the uncanny valley — is both unsettling and a little thrilling.
It makes sense, then, that we love the screwed-up face swap — those images where the program finds a face in an electrical socket or a stove burner — even more than the perfect one. On the one hand, it’s unnerving to see our faces treated as collections of pixels interchangeable with a variety of objects. And the objects the app tends to mistake for faces are telling: If an artist superimposed an electrical socket on someone’s face we’d call it clichéd; when a machine does it, it’s more than a little discomfiting.
But on the other, it can be a relief to know that, even though code is just a complex set of step-by-step instructions, it is still in some way fallible. It cannot yet fully approximate human technique and appearance.
At least, not yet.