Algorithms themselves are merely mathematical shortcuts to solving problems. Inherently, they are neutral in ethics, as they are merely a tool to decide what information is most relevant to which users. Perhaps the most controversial quality of algorithms is their capacity to influence a large number of people, as the control over who gets to decide what information gets pushed ultimately decides the thought of the masses.
Though this infrastructure has the capability to do immense good (pushing domestic violence hotlines to those the data networks suspect are in dangerous situations or promoting accurate COVID-19 information), it also has the potential to do great evil (spread mass misinformation and sway public elections).
In popular dialogue, algorithms are mostly discussed within the context of social media. These algorithms decide, according to their supposed perfected calculations of relevancy, what media to place at a higher priority on ones’ feed. Nonetheless, certain random videos, seemingly without reason, get pushed to millions of viewers, while other videos are hidden. The implication is dangerous; misinformative or dangerous videos have the capability to reach millions of views while informative and factual videos can be pushed to the bottom of the algorithms’ priority list.
The decision of what information gets emphasized is not up to experts in the field or even to viewers themselves. Rather, it is up to a mathematical algorithm that is created by humans, and will thus be a mathematical sequence of codified implicit (or perhaps even explicit) biases. For example, Amazon’s 2014 hiring algorithmic tool codified a bias against female engineers by learning to discriminate against certain women’s organizations and colleges listed on resumes.
The question, therefore, becomes, who should be allowed to regulate these all-encompassing algorithms? Legislators are so out of touch with technology that they are incapable of conceptualizing algorithms, let alone legislating them. In perhaps the most disappointing long-awaited testimony of Mark Zuckerberg, senators like Senator Bill Nelson (D-FL) and Senator Orrin Hatch (R-UT) spent precious minutes, respectively, questioning Zuckerberg about how targeted ads worked and how Facebook’s subscription-less business model was profitable, rather than digging into the real issue of privacy and regulation.
Not all hope is lost, however. Representative Alexandria Ocasio Cortez, meanwhile, very sharply and knowledgeably grilled the mogul about the Cambridge Analytica data scandal that compromised the data of millions to political advertisers.
Nonetheless, the question regarding whether Congress, many of whom were evidently unable to grasp Facebook’s business model, is equipped to legislate effectively on the matter remains. As Congress continues to be indecisive and unable to pass any federal legislation regulating big data to the extent that California as a state has with its 2018 California Consumer Privacy Act, the reach of these algorithms grows in tandem, their power unchecked.