This Halloween, a Social Experiment Will Allow Internet Users to Control the Actions of a Real Person

MIT Media Laboratory’s BeeMe is the love child of ‘Black Mirror’ and psychologist Stanley Milgram’s notorious experiments on free will and obedience

BeeMe:MIT Media Lab.jpg
Celebrate the spookiest night of the year by participating in a "game" that lets you guide a real person's actions BeeMe/MIT Media Lab

What do you get when you combine Netflix’s “Black Mirror” and social psychologist Stanley Milgram’s controversial electric-shock experiments on free will versus obedience? No one can say for sure, but it's likely this dystopian love-child would look very much like BeeMe, an online social experiment starting at 11 p.m. on Wednesday night.

As Dave Mosher reports for Business Insider, the folks at Massachusetts Institute of Technology’s Media Laboratory are behind the experiment, which will allow a group of internet users to control a real-world individual as he or she attempts to defeat an evil artificial intelligence called Zookd.

“In times where algorithms make most of our decisions for us, one individual will entirely give up their free will for a day, to be guided by a large crowd of users through an epic quest to defeat an evil AI,” the BeeMe website states. “Who is in charge? Who is responsible for one's actions? Where does the individual end and others begin?”

The MIT lab has something of a tradition of marking Halloween in such a frightful fashion. In years past, they’ve brought us the Nightmare Machine and Shelley. The former transformed normal photographs into nightmare-inducing scenes, while the latter penned horror stories in collaboration with Twitter users.

But this year, the team has really upped its game. According to the project’s description, BeeMe is the world’s first “reality augmented game”—a playful reversal on the burgeoning field of augmented reality.

Niccolò Pescetelli, a collective intelligence researcher at the lab, tells Mosher the game will feature a trained actor hired to portray the human guinea pig at the command of a captive online crowd. The actor’s location and identity won’t be revealed, but participants will be able to watch his or her actions through a voyeuristic video recording.

To direct the “character”’s steps, internet users must submit commands that can range from the mundane decision of “making coffee” to the more foreboding order of “running away.” Participants will vote on the assortment of submitted actions, and the actor will perform those that field the most votes.

There are, of course, parameters to the game: BeeMe will stop short of allowing commands that violate the law or place the actor, their privacy or their image in danger. But the game will not impose limits beyond that, which leaves the door open for plenty of mischief.

What are the implications of erasing someone’s free will, even for just one night with set limitations? Back in the 1960s, Milgram’s studies shed some light on this, finding that individuals under the influence of an authoritative figure often follow orders to an unprecedented extent. In his experiments, 65 percent of participants acting as “teachers” agreed to administer a maximum 450-volt shock to unwitting “learners,” who were actually actors simply pretending to experience the effects of the deadly shocks.

As the Verge’s Shannon Liao points out, BeeMe feels almost as though it was lifted from a 2014 “Black Mirror” episode entitled “White Christmas.” In the show, a skeezy dating guru outfits his subjects with a chip that enables him to see and hear what his protégées experience. As this is “Black Mirror,” the episode predictably shows how technology brings out the worst of human nature.

BeeMe, of course, isn’t quite on the level of “White Christmas,” but Pescetelli of MIT tells the Boston Globe’s Steve Annear he expects “a bit of online trolling to try to tip the narrative and get the character to do outlandish things.” Still, Pescetelli emphasizes the precautions the team has put in place for the actor’s safety.

“BeeMe will redefine the way in which we understand social interactions online and in real life,” writes the MIT Lab in an unsettling statement, which predicts the experiment “will push crowdsourcing and collective intelligence to the extreme to see where it breaks down."

Pescetelli tells Annear that in addition to sparking some Halloween fun, the team hopes to see if a large group is capable of making one individual execute a fluid series of tasks or if the information overload will devolve into a series of disjointed antics.

A tweet by BeeMe’s account back in August points to one strategy for success. Quoting Charles Darwin, the post writes: “[In]...the long history of humankind (and animal kind, too) those who learned to collaborate and improvise most effectively have prevailed.”

Get the latest stories in your inbox every weekday.