It’s more than halfway through 2018, and the news is still flooded with stories from the 2016 American presidential campaign. It’s no secret that the campaign was marked with a particularly high level of emotion, giving rise to many theories, some founded, others not. When the news broke that Cambridge Analytica had used ill-gotten user information from Facebook in order to target voters, it became very clear that we were not ready for role technology can play in our elections. Machine Learning has been rapidly implemented in many industries before this, even social media, but somehow, better picture tagging didn’t translate to the real-world possibilities. Since 2016, people have been wondering what we can do to prepare for this in the future. Whatever you thought the solution may be, you probably didn’t expect this: you can (hopefully in the future) play Machine Learning President.
In the wake of the combination of money and technology influencing an election on an unprecedented scale, Brett Horvath and Berit Anderson, who co-founded Scout AI, created the game Machine Learning President. Inspired by what they described as “an invisible machine that preys on the personalities of individual voters to create large shifts in public opinion,” they crafted a game to get pro-democracy groups thinking about how money and technology played a part in the 2016 election, as well as to help them prepare for the same in 2020.
In Machine Learning President, players take on roles that are intrinsic to an American presidential election: candidates from both major parties go through a primary election, then an actual election, while special interest groups, powerful donors, and other characters try to push their policies and sway the public. The goal is to educate the players about the role AI played in the 2016 election and is likely to continue playing in the future, to protect democracy. Currently, due to lack of understanding and therefore lack of intervention, things as simple as something you reacted to on Facebook can give third parties enough insight to feed you enough targeted information to sway your opinion on major issues. The hope is that with people aware of this phenomenon, it can be minimized and, when occurring, less successful.
So what do you do to avoid finding out in the future that “your favorite politics page or group on Facebook didn’t actually have any other human members, but was filled with dozens or hundreds of bots that made you feel at home and your opinions validated”? Well, you certainly can’t stop innovation, and as long as a thing exists, someone will misuse it. Whether or not playing an elaborate strategy game will help the public at large avoid potential manipulation, it certainly shines a light on a problem that largely remains in the shadows: most of us aren’t aware enough of what technology can do. As in most scenarios, the best thing to do is get educated about the subject, and if that can be done by playing a game, well, all the better.