The Algorithms Behind Social Media
While many senators expressed concern during the testimony of Facebook whistleblower Frances Haugen, the key question remains: What are they going to do about it? Wherever people land on the truths or misconceptions Haugen raised is not the key issue. The real threat that was exposed is how social media platforms are data regulation and designing algorithms that unknown to followers, are influencing and changing human behavior. Without regulating this core issue, dangerous practices will continue.
The initial promise of social media technology, as a way to stay connected to each other, build bridges and move humanity forward, has a darker side which is finally coming to light and fueled by the desire to increase engagement, and thus profits, at all cost. To address the dire consequences social media can cause, Haugen and others are advocating for greater transparency and stricter oversight. And, while nothing has been announced yet, senators are now talking about the possibility of strengthening privacy and child protection regulations, and applying anti-trust laws.
While well-intentioned, these efforts are wholly inadequate and don’t go far enough. Instead, Congress must put an end to the use of algorithms to influence or change human behavior – period. These algorithms, designed to increase engagement, advertising, and ultimately profits, end up fanning the flames of controversy, conspiracies, and disinformation that influence people’s emotions and actions.
The power to change human behavior
Studies have shown how social media sites can affect the body image and moods of children and teens, leading to increased anxiety, depression, and suicidal thoughts. And, we’ve all seen the increase of polarization in politics, vaccines and, other issues and how this has led to confusion, anger, and in extreme cases, violence.
Concerns about using hidden techniques to influence human behavior is nothing new, nor is the outrage that comes along with it. Subliminal advertising messages, such as those flashed on movie screens telling people in theatres to “eat popcorn,” have been banned in many countries. While the jury is still out as to whether those techniques are effective, the FCC and others recognized their potential for abuse.
Yet, punishing, regulating or breaking up Facebook won’t solve the problem. Facebook and other social media sites did not intentionally try to create these issues, and to turn it into some evil plot designed to destroy humanity is best left to Hollywood. Rather, they’ve been doing what all companies do: working to increase profits. Unfortunately, the technique they use to increase engagement, advertising dollars, and ultimately profits, has inadvertently led to large social, emotional and political problems.
Unless stopped in its tracks, the practice of using AI algorithms to manipulate people will continue to grow across platforms because, well quite simply, it works.
Addressing the problem at its source
It’s not realistic for social media platforms to police themselves and balance shareholder profit against the negative consequences that might ensue from their practices. Nor is it enough to try to address or data regulation the symptoms that have arisen from manipulative practices, which is where Congress seems to setting its sights.
We must go further and ensure there is a strong regulatory body that can set strict policies to end the use of algorithms that can influence human emotion and behavior. And we need to give this agency the teeth to enforce it, by banning behavior-modifying algorithms outright, and enacting severe criminal and civil punishments for companies that violate that ban. To ensure compliance, companies must become more transparent, and open up their data regulation and practices to the regulatory commission and other relevant bodies through ongoing inspections and audits.
We certainly have a way to go in regulating social media and it’s no easy task. Yet, until lawmakers address it, they will just be dealing with the many different ways that human manipulation algorithms can be manifested, and get caught up in a fruitless effort to try to stamp them out in a whack-a-mole scenario. We must establish broad-based algorithm manipulation laws before it’s too late.
Sign up for our Monthly Newsletter: