The Uncanny Valley of Social Media


How does reading that make you feel? Good? Bad? Pleased? Uneasy? Physically ill? Some young users see the word “Facebook” and feel revulsion. Their trust as users has eroded.

I'll take us through Facebook’s past to the evolving fields of animation, robotics, and virtual reality as I attempt to address a common question: why does using Facebook feel so unnatural?

But I’ll start by answering this question: did it always?

Famously created in a Harvard dorm room, Facebook has grown to be the dominant medium for online social interaction among young American adults. At its inception, Facebook focused on serving the isolated Harvard community. If you joined the site and added a friend, chances were strong you already knew the other person by name or face and that you had multiple mutual “strong-tie” connections.

A dramatic rendition of a Harvard dorm room.

A dramatic rendition of a Harvard dorm room.

As Facebook slowly expanded, it allowed other college students in the Boston area to participate in its network. Then Ivy League and Stanford students were invited. From these beginnings Facebook grew to the behemoth it is today, now having over 1 billion monthly active users and a strengthening mobile presence.

Demand for Facebook’s growth also meant demand for Facebook to change. In one of the original versions of the site, users simply maintained a personal page that friends could discover and interact with. As features were added, users at first delighted in their new capabilities. Suddenly you could post on walls, invite friends to events, and create groups. Now we can browse company pages and news feeds, play games hosted on the Facebook platform, and yes, still “poke.”

Somewhere along the line, though, we began to feel betrayed.

Multiple placed ads interfere with our engagement or disquiet us with their specificity. Fake profiles and spammers demand our attention. And recently, users were outraged by Facebook’s willingness – perhaps eagerness – to manipulate their emotions by altering the content of their news feeds.

We used to love the freedom Facebook afforded us in the online social experience. We used the service gladly, and so did all our friends. We had nothing but good will. Facebook simply wanted to approximate and augment our real-life social interactions so that we could take our existing relationships into the evolving online world.

So why did it all go wrong?

The answer is simple – it had to.

What do I mean? To answer this, let’s turn away from social media and toward the fields of robotics and animation.

As our technological capabilities have increased over time, our ability to approximate the human form using computers has grown dramatically. Robots have evolved from simplistic humanoids to near-lifelike representations of actual humans. Animations, once grounded in the two-dimensional plane of cartoons, have grown so complex and commonplace as to appear in every summer blockbuster. Visual and digital effects artists earn – or should earn – top billing alongside the stars.

So if technology has come so far, why are we still so repulsed by videos like this?

In 1970, the Japanese roboticist Masahiro Mori coined the phrase “uncanny valley” to explain our complex empathic response to human-like robots. His hypothesis was that a human observer’s emotional response to a robot was closely related to the robot’s appearance: as it grew more human-like, observers became more positive and empathic towards it.

However, Mori also explained the existence of a critical point beyond which observers were actually strongly repelled by the appearance of the robot. As robots come to resemble a human almost perfectly, a feeling of not-quite-rightness overcomes the observer and causes a powerful negative reaction. The feeling is so strong that any productive interaction between the human and robot is eliminated.

This is the uncanny valley – the chasm separating useful tools of human representation and actual humans. And in this valley, we feel lost.

The Uncanny Valley - credit in image

The Uncanny Valley - credit in image

Let’s climb out of this valley and return to Facebook.

Recall how the service has evolved. At first your Facebook profile was an imperfect approximation of your social identity. A static photograph of your face and a few lines of text summarizing your beliefs served to encapsulate your existence. No one could confuse your Facebook profile with your actual social persona – the complexity required to represent it did not exist.

As features were added over time, our profiles became part of our identities. We posted photo albums, added “life events,” checked in at landmarks, and shared deeply personal status updates. Instead of simply mirroring our rich social lives on Facebook, we lived fully online. Our online interactions were no longer merely reflections of our real-world friendships. We used them to create our perceptions of ourselves and others. Our digital selves became extensions of our physical selves.

So as our personas and our technologies have evolved – as Facebook’s services evolve to allow for more complex online interactions – we have gained the ability to nearly approximate our real-world relationships on the Internet, such that we can almost seamlessly flow between our online and physical worlds.

But it’s not quite perfect. And there is the problem.

My perception is that we are in the uncanny valley of social interaction online. We’re close to something good – close enough to know that we are close. But also close enough to know that something is terribly wrong.

There is still something about the way we conduct ourselves online – whether our platform is Facebook, Snapchat, Twitter, forums, comment sections, and so on – that is perverting our social interactions and our perceptions of each other and ourselves. These services allow us to be human-like, sharing our ideas and faces and actions, but we are still forced to be distinctly non-human in a way that repulses us. All this is happening while we struggle to learn how to absorb our online personas into our identities.

Let’s be clear. The problem isn’t the users. It’s not the fault of a particular generation or group of narcissistsThe problem is the systems themselves – not just in the way they’ve been commercialized, but also in the way they’ve attempted to accurately replicate, augment, and replace our social interactions while integrating new aspects of our digital personalities.

As investors demand growth and users demand new services, our social platforms try to do too much. In pursuing the essence of our social interactions, the platforms are coming closer, but they’re still not quite right.

And in the meantime, we won’t feel quite right either.