Silicon Valley's Futile Search for Utopia Via the 'Perfect Algorithm'
I find myself thinking quite a bit about algorithms these days. The word itself derives from the Latinized name of the Persian scholar Mohammed ibn-Musa al-Khwarizmi, reputedly the inventor of algebra, who flourished during the Abbasid Caliphate in ninth-century Baghdad. 1200 years later with the advent of the computer, the age of programming and machine-learning dawned, refining and applying al-Kwarizmi’s scheme of numeration into conceptual regions he could never have imagined. According to the standard definition, an algorithm is a set of rules determining the nature and order of computer calculations, the major component of search engines that canvass data bases cued by key words.
In The Master Algorithm, Pedro Domingos writes: “A programmer—someone who creates algorithms and codes them up -- is a minor god, creating universes at will. You could even say that the God of Genesis himself is a programmer.” The serpent in Algorithm Eden is complexity -- of space, time and human limitation -- creating a world that grows “increasingly fragile.” Data is turned into information and information is transformed into knowledge.
But the situation grows increasingly tangled when we realize that algorithms can reflect a programmer’s ignorance or prejudice or explicit design, and that algorithms can also learn to rewrite themselves, that is, they can also be self-programming, introducing a degree of uncertainty into the original parameters. Knowledge may be skewed, infected by error, and even prey to delusions -- a tree whose fruit should not be plucked and eaten.
Similarly, Frank Pasquale in The Black Box Society: The Secret Algorithms that Control Money and Information alerts us to the troubling fact that algorithms, often opaque to their own programmers, can serve to reinforce social taboos, prejudices and prior assumptions which reflect the unconscious attitudes of the programmers. But these attitudes may also be quite conscious, introducing a propagandistic element into the algorithm. Pasquale writes: “The proprietary algorithms… are immune from scrutiny,” rendering us vulnerable to surveillance, censorship, and coercion masking as persuasion, and so “undermining the openness of our society.” He points out that mining data from social media in the hunt for potential malefactors “comes with a high risk of false positives.” We should be aware, too, that it comes with a growing certainty of false negatives. Indeed, given the monopolistic power of the major social media networks, practically all promoting progressivist memes and left-wing politics, such contamination is inevitable.
The principal social networks -- Facebook, Google, Twitter, YouTube, Patreon, etc. -- all rely on secret algorithms derived from human emotional and ideological input. As Jim Treacher writes at PJ Media, “Tech companies are notorious for their liberal culture. … Worse, tech companies like Facebook, Google, Amazon, and Twitter have relied on the Southern Poverty Law Center (SPLC), a far-left smear factory that brands Conservative and Christian organizations ‘hate groups.’” Clearly, these networks are not simply digital common carriers but a species of political cabal.