As AI-powered software that can identify human emotions becomes more commonplace, a new browser games wants to illustrate the limits of the technology. Spotted by , the Emojify Project was created by a led by University of Cambridge professor Alexa Hagerty. You’ll find it over on . It will ask you to look at your computer’s web camera and try to produce six different emotions: happiness, sadness, fear, surprise, disgust and anger. As you play the game, what you’ll notice is that it’s easy to fool the software. For example, you can fake a smile to trick it into thinking that you’re happy. Ultimately, that’s the point of the experience.
What the game is trying to get at is that there’s often a disconnect between the emotions we experience internally and the face we show to the world. For example, think of when someone cries. In most situations, they’ll do so because they’re sad, but people are also known to shed tears of joy. Humans are messy and complicated, and we all experience emotions differently. Suggesting it’s possible to distill how we feel in a given moment into a simple set of emojis does none of us a favor. As with , there’s also the problem of racial bias. A found that programs like Face++ will typically rate the faces of Black men as angrier and unhappier than their white counterparts, no matter the actual expression on display.
And yet the technology is becoming increasingly common. It’s not just something you see at the anymore. Organizations and governments are using it for everything from screening to passport security. Even Apple in the space at one point. In showing the flaws of emotion recognition software, the project hopes to spark conversation about its use. “Our aim is to promote public understanding of these technologies and citizen involvement in their development and use,” the group says. “We believe that through collective intelligence and sharing perspectives on such important issues, we can empower communities to promote a just and equitable society.”