Marxism and Data Science

There are benefits that data-driven automation has given to humans, but at what cost? Can we blindly seek value without the aid of a human perspective guided by philosophy? Let’s consider Karl Marx’s writings on capital and machinery. Marx’s Prophecy states that at the breaking point of human misery, a communist revolution will establish a dictatorship of the proletariat. As the market becomes more concentrated and companies absorb others, jobs disappear and a permanently unemployed class is created. 

Workers bear the costs of increased productivity generated by automation. AI bosses have created an Orwellian work environment in many industries. AI scheduling systems leave workers always on call, always at capital’s disposal. AI overseers identify when workers are “not working”, as the WorkSmart software takes pictures on the workers’ webcam every ten minutes and watches their open applications and keystrokes, causing them to avoid using the bathroom or even taking a break to think. Amazon workers and meat packers are forced to keep up with machines’ inhuman speeds, putting them in danger of fatal or debilitating injury. 

Outside of the workplace, data science is present in our personal lives--particularly social media. Autonomism, a theory based on Marx’s Fragment on Machines, explores how capital relies increasingly on accumulated science and knowledge, and capitalist relations have expanded beyond the workplace to encompass all social relations. The internet is a machine which extracts value out of individuals’ web activity and content, in exchange for nothing in return from the companies that own the sites and generate “big data” from their users. The individual invests their personality and emotional attachment, but becomes alienated from those. As advertisers pay to access Facebook, the users have become the product. More and more of our lives are being replaced with this privately owned technology. On platforms like Facebook, we must conform to their automated enforcement of their terms of service. Because Facebook has taken the place of public services like the mail, we must ask what rights Facebook has to limit our speech, as Facebook represents private interests. Facebook has censored anti-fascist and anarchist groups, and has conducted emotional manipulation experiments on the site without users’ consent. Meanwhile, the AI industry tends towards monopolization, as billions have been spent on buying companies with AI technologies and on accumulating more data. Unregulated capitalism flows power up the pyramid to an increasingly smaller group of people. Regulations cannot keep up with technology, resulting in an increasingly deregulated landscape. Who profits and how? Philosophers (and data scientists) must ask: Is data labor? How might this labor be repaid? 

Philosophy can allow data scientists to predict how much of humanity can be replaced by machines in the future. Monism states that humans are composed of inanimate substances and thus machines are capable of taking over every possible human job. Meanwhile, Dualism claims that a soul separates humans from objects and thus, some jobs cannot be replaced as machines cannot replicate empathy, creativity, or critical thinking. Large-scale worker replacement is inevitable, and it is up to us to decide how to respond.

There are different viewpoints on what can be done. Proponents of the Free Market argue that as technology causes costs to decrease, consumers will be empowered to spend more on luxuries, driving up demand and creating more jobs. Meanwhile, Neo-Marxists suggest that privately owned means of production must be handed over to “social ownership of machinery, land, and resources”. Data science can allow the creation of a planned economy, allocating resources according to big data analysis and robust feedback loops. Somewhere in the middle, government intervention such as UBI could suffice, as many countries have adopted socialist policies without becoming communist. 

How much intervention should the government take? What existing models (such as Norway) could we take inspiration from? What regulations are needed to protect humanity? Can AI be used for good in a free market? These are just a few questions that data scientists must ask, and this is why data science requires the social sciences to predict the human impacts of the technologies they create. To quote Jurassic Park: “Your scientists were so preoccupied with whether they could, they didn't stop to think if they should.”