Increasingly – and often by stealth as they work inside your innocent-looking search engine – algorithms have entered not just our computers and working lives. They also govern more and more eventualities of life in so-called advanced societies and elsewhere.
The existence of algorithms might be a sign of civilization, but it might also be a sign of what French philosopher Foucault calls madness. As human decision-making is handed over to machines, these machines can make rather irrational, discriminatory, and outright mad decisions.
Long gone is the time when algorithms were only used in mathematics, IT, software, and computer science. Historically, the word “algorithm” is an Arabic term meaning a decimal notation of numbers, while in Greek, arithmos is simply a number. Our modern-day arithmos or algorithm came to us via Latin’s algorismus.
In any case, an algorithm is a finite sequence of rigorous instructions. These are used to solve a class of specific problems. Algorithms can also be used to perform computations with feedback loops. Often, they are applied to complete calculations and data processing. More advanced algorithms can even execute automated reasoning.
They are also used in mathematical and logical tests for corporate recruitment purposes. None other than Alan Turing relied on the human-like characteristics as descriptors of a machine that cracked the Nazi code. Later he was driven to suicide by the English establishment, after, of course, he had deciphered the Enigma code, saving thousands of lives.
One of the best and most current definitions of algorithms comes from Cathy O’Neil, who noted that algorithms are opinions put into a mathematical formula. In other words, there is a person who sets up the algorithm for a specific purpose.
It remains imperative to remember that virtually all – rather scientific-looking – algorithms are plagued by hidden interests – often the interest of the users who usually asked an algorithm to be created, e.g., a company like Amazon. In other words, in many cases it is capitalism that drives the design of algorithms.
As a consequence of all this, algorithms increasingly sidestep the role of the subject and even human subjectivity when constructing knowledge, the automatic data analysis based on a mathematical formula. This chisels away the humanist project of the self. Philosophy calls this personhood.
In short, the application of algorithms weakens our ability to form and articulate what we think (e.g., search engine algorithms), what we like (Facebook algorithms), and what we want (Amazon’s algorithms).
Worse, this version of algorithmic knowledge destabilizes a core process in the construction of subjectivity. It eliminates self-reflection by deducing – or even eliminating – the active participation of the individual in the creation of knowledge. Beyond that, algorithms also endeavor to create positivistic and even behavior-manipulating knowledge. Algorithms can eradicate nearly all self-reflective and communicative facets of knowledge creation.
As algorithms are becoming more sophisticated – and are spiced up by artificial intelligence – they have the true potential of becoming independent agents in the formation of our social – and more importantly – our democratic and political society.
It may well be almost inevitable that algorithmic knowledge will attempt to compete with and potentially even bypass, a particular and above all – still! – the unique aspect of being human: our individuality, agency, and subjectivity.
Most deceptively, the apostles of algorithm assure the user that algorithms will grow the realm of personal freedom as these systems offer a richer, much more truthful, and much more precise form of knowledge. The positivist ideology of neutrality is always a helpful vehicle when corporate interests need to be camouflaged through engineering-like “techno-solutionism” – they misbelieve that all problems have a technical solution.
Essentially algorithms contain the deeply ideological promise to make our workplace and society freer, filled with more emancipated human beings. In reality, algorithms can very easily do the exact opposite. Algorithms – mostly automatically – order data and things. They create a new, and most importantly automated, order of things – the order of the automatic algorithm. Potentially, this might be even worse than what French philosopher Foucault outlined in his seminal masterpiece “The Order of Things”.
In other words, algorithms establish – if not cement and solidify – an existing social and economic order. For algorithms to work, it requires no subjects at all. Instead, human subjects are turned into objects of power, as outlined by the Polish-British philosopher Bauman, though they do not live up to Wittgenstein’s dream of leaving everything as it is.
Wittgenstein may not like it, but algorithmic computer systems do constitute an entirely new regime of knowledge based on opinionsturned into mathematical equations. And these new systems have – and will continue to have – a huge impact on contemporary work, management as well as social and, of course, consumer life. While gaining increasing importance, the world of algorithms remains largely unregulated and exists outside of the domain of democratically assured regulation.
Almost inherently, many – if not most – algorithms are, in fact, future-oriented. They are designed to predict future behavior and worse, control and manipulate human behavior. One of the most sophisticated uses of algorithms can be found in Amazon. Some might be tempted to claim that Amazon knows that you will buy condoms or diapers before you do!
To do that, the people behind algorithms need to construct a particular type of a human being. This is a human being that is predictable – and hopefully controllable. In turn, these data produced by human beings and collected, analyzed, and used against human beings can be used to refine the algorithm further.
This is done in the hope that all this will lead to, among other things, improved crime fighting, euphemistically known as predictive policing, that often replicates entrenched prejudice. Algorithms are also used to improve corporate profits, as in the case of Amazon.
For algorithms to deliver on this, corporations like Amazon need to know who we are, where we live, what we like and don’t like, and what we want and buy. Beyond that, algorithms also like to assume a dormant subjectivity, a subject that is really more of an object – the hidden object inside us – that can be commercialized and turned into something useful to corporations.
Such an individual becomes structured within the contours of a mathematical formula that is used in an algorithmic environment. Worse, such an individual whose behavior can be predicted, measured, controlled, and ideally manipulated can rather easily become an Uber-engineered human being. This algorithm-driven individual is imagined to be radically different from the way an individual was once imagined under modernity.
Since the people who set up algorithms create a kind of knowledge-making machine, they see human reality in a very particular way. This reality is radically different from the who and what we are today. This represents an entirely new Malaise of Modernity.
Of course, such algorithms have already become a new form of a kind of non-democratic – and, if you will, corporate-driven – governing agent that predicts, manages, controls, and even manipulates human life, consumers, and perhaps even entire populations.
This can be done without the need for any form of what might be called knowledge of critical subjectivity. This is what German philosopher Adorno calls the mündige individual, a self-reflective, self-critical, mature, and autonomously acting individual.
Almost necessarily, algorithms need to work behind the backs of those “for” (or better “against”) whom they are designed. These systems remain opaque while their invisible script runs behind a colorful computer screen.
Meanwhile, their ramifications are increasingly apparent to all of us. Algorithms already make not just an aesthetic judgment for you but also hiring decisions for the human resource manager, among others.
Perhaps even more damaging is the fact that algorithms make those decisions not “with” you but “for” or, in many cases, “against” you. Consequently, algorithms promote a mode of non-communicative knowledge. This is the utter destruction of what German philosopher Habermas calls communicative action and ideal speech.
A world where Habermas’ ideal speech hardly exists but which is increasingly governed by algorithms is the world of corporate management. Overall, one might say that the level of algorithmic management differs from company to company, perhaps with bottle-peeing Amazon on the “full-scale” end of the spectrum.
While it might be hard to predict where management is going even in the near future, one might like to argue that management’s self-interest is not to move to level 6 of full AlgoM. This might render a large section – or even management as such – obsolete.
And for that, top management might replace middle management and workers by moving towards full AlgoM. At this level, algorithms define virtually all performance, evaluation, and control functions “without” the involvement of real managers.
Whether in management as AlgoM, at the police, or in corporations such as Amazon, algorithms offer a radically new and very different way of knowing, understanding, shaping, and manipulating the world and – worse – the individual.
Algorithms have the power to change the very foundations of what it means to know, to understand, and to reach decisions. Yet, it remains imperative that these algorithms and the purpose-built knowledge they create are not solely understood in mathematical, engineering, and technical terms. As opinions are placed in mathematical formulas, algorithms increasingly underwrite many operations in business and in society.
Algorithms provide a very different conception of what it means to be human. Algorithms break the human-knowledge link as they seek – and, in fact, do – create knowledge that is gradually disconnected from subjective individuals.
As a consequence of all this, algorithms need to be understood not just as one of the most recent, but also as perhaps the most severe threat to human life as we know it.