AI: The machine intelligence of imperialism – Part 2

Part 1 discussed “Digital labor and material.” Part 2 takes up how capitalism uses Artificial Intelligence to push workers to the limit, some of which I experienced while working at the Tesla Gigafactory. 

My work at Tesla provided one example of “surveillance capitalism,” a term coined by Shoshana Zuboff, a Harvard professor who studies and writes about the sociology of technology. An October 2019 article in The Guardian reporting on Zuboff’s tour discussing her book on surveillance capitalism reports on the situation we workers face and the term Zuboff coined:

She said:  “It is a movement founded on predictive algorithms, mathematical calculations of human behavior. Surveillance capitalists ‘sell certainty to business customers who would like to know with certainty what we do. Targeted adverts, yes, but also businesses want to know whether to sell us a mortgage, insurance, what to charge us, do we drive safely? They want to know the maximum they can extract from us in an exchange. They want to know how we will behave in order to know how to best intervene in our behavior.

“The best way to make your predictions desirable to customers is to ensure they come true: ‘to tune and herd and shape and push us in the direction that creates the highest probability of their business success.’ There’s no way ‘to dress this up as anything but behavioral modification.’ In 2012 and 2013, Facebook conducted ‘massive-scale contagion experiments’ to see if they could ‘affect real-world emotions and behavior, in ways that bypassed user awareness.’”

Responding to Zuboff’s point, I’d say the ruling class does use technology to manipulate us and also offers us minor amenities — like “free” Spotify for music at work or “free” trail mix in the breakroom. These crumbs are meant to distract us from the ways in which digital technology is being used to enhance the exploitation of the working class. Disguising exploitation increases antagonisms among the workers, with their physical workplace itself and with their supervisors, all in a prison-esque atmosphere.

From architecture to information systems, the modern computer office space holds a semi-carceral status in its design. This prison-like status goes beyond the work zones. Capital, that is, the boss, using digital technology, encroaches on rest areas, common spaces, even public and domestic spaces.

While the term “surveillance capital” is colloquially useful, it is my belief that relying on this term too heavily creates an artificial distinction that can lead us to believe that some corporations, or some bosses, are better to work for than others, that it is up to their individual ethics how intensely the workers are exploited.

Capitalists aim to maximize profits

The truth is that capitalists do what is in their fiduciary interests. That is, they are driven to maximize profits, and that means maximizing exploitation of labor. At the end of the day, every boss is a class enemy of the workers and will use whatever tools are necessary to continue the exploitation of the working class, including superexploiting workers of oppressed nationalities, and will plunder the land and its resources.

Furthermore, this technology measuring labor productivity is no crystal ball. Rather it exists as a logistical and statistical analysis that predicts the probability that something could happen — not a certainty. It employs strategies, interventions to maximize this probability through behavioral modification, when an individual is making a specific decision (such as ordering food online) based upon previous behavior patterns.

While working at the Tesla Gigafactory, I experienced these interventions.

Our office was designed very much like a cross between a military installation and a prison. There were very few windows, and the windows we did have all had thick steel bars. We had security gates and constant facility monitoring, all focused on the workers.

Despite the heavy surveillance we faced, when it came to protecting the vehicles owned by workers, the plant security staff was unable to protect these vehicles from being broken into or stolen. And when this happened, security prevented the workers from taking photos for insurance, claiming this was “a potential breach of the NDA and Tesla security.”

Specialized algorithms tracked our keystrokes and used them to generate numbers that measured our performance. Management graded us harshly using these metrics. Although I had several documented disabilities that should have prevented harsh treatment, supervisors lectured me about my productivity several times. Why? Because my tracker showed too many five-minute spaces in my day when I needed to pee or massage away my wrist pain — caused by carpal tunnel syndrome combined with fibromyalgia (generalized widespread pain and chronic fatigue).

Our office was extremely dark, while the keyboard monitors were too bright, which hurt my eyes to look at —  especially for hours on end. The tracking forced team members to compete with each other through the quality check system. The feedback did not improve the self-driving AI we were working on — but instead was used to determine who would receive a raise at the end of the half-year term of an amount ranging from one to 10 cents.

Since I was a union organizer, I found that management increased their monitoring of me to include constantly surveilling me on my breaks; and at several points I remember my fellow union committee members and I needing to go for walks outside the facility, because management was recording our conversations with other workers to use against us.

Surveillance to disrupt organizing

Our cellphones were also jammed except for emergency 911 calls. This made communication with each other or with the outside world extremely hard without encrypted communications, specifically chat applications that ran on the WiFi. If we used the company’s internal digital forum, conversations in one “private” area of a digital chat space or in a “private” email thread would suddenly become open topics of conversation between a supervisor and a department. Thus the need for real-time and truly private communication was born.

Tesla workers are not alone in being subject to algorithmic management. Workers of all types — ranging from restaurant service staff to delivery drivers, medical staff, engineers, designers and insurance adjusters — are subjected to it in some capacity or another.

This increasing growth of “digital managers” shows us that even those who are not direct employees, but might consider themselves small business people, have interests threatened by the corporation’s digital technology. The sooner we all recognize who our enemies are — the big capitalists who exploit all of us — the more likely we can make our working-class interests prevail.

Tech industry management gains if it tricks us into believing that we can be pro-establishment techies who sell their lives to some firm and in exchange can own our personal mini-tech business or be an “influencer.” This is a fantasy which is described in more detail in an article by Jathan Sadowski.

Discussing a concept called “the mega-algorithm,” Sadowski describes this myth: “The mega-algorithm doubles back and promises your reunification with this alienated self through ‘authentic’ (or ‘creative’ or ‘social’) work completed on your own time. … The techno-utopic discourses of emancipation and community that surround the technologies and sociopolitics of that underpaid laborer trying to hustle for a paycheck. No you’re actually an entrepreneurial individual,” Sadowski adds with sarcasm, “building your personal brand.” (New Inquiry, April 2015)

To put this differently, our networked and computerized systems exist as a sort of massive digital ecosystem of machines that work in tandem with each other (hence “mega” + “algorithm”). These machines require some type of human labor to operate correctly, and the ruling class exploits the workers who keep these machines working. If we recognize our status is limited to being the tiny cogs in a great exploitation machine, we become dangerous to ruling-class interests.

To prevent this change in our consciousness and keep us complacent, the capitalist bosses, or people they hire, tell us that our hard work is instead part of “building your own brand” or some other capitalist myth meant to make our exploitation seem worthwhile and sensible to ourselves. This general pattern holds whether it involves maintaining the highest tier machine learning platform to the most basic internet routing system.

As I will take up in future articles, there are workers behind every single tool and machine in existence, and it is the responsibility of these workers to recognize their class interests in the global class war. A first step is to withhold their labor from militarized or quasi-militarized technology. For the revolutionary and progressive members of our class who are outside of technical work, we must remember to stand in solidarity with one another and support our fellow workers.

Part 3 to come: The media of digital capital.