Computer says no

Faceless capitalism and the normalisation of algorithmic culture

Call centre blues

My first job was at a call centre for a telephone company, back in the late 1990s. This company, recently privatised at the time, was undergoing the closure of its client-facing offices, transitioning all customer services to a single phone number. The publicly announced rationale was that, being a telephone company, it made sense to cater to all customer needs over the phone.

My job was gruelling, badly-paid, and precarious. We’d take hundreds of calls a day, with just a seven seconds’ gap between callers. We were granted just two fifteen-minute breaks and thirty minutes for lunch. We were supposed to record an average contact time of three minutes per call at the end of the shift. This was a perverse requirement almost no one met, and that our managers would use to threaten us with all sorts of punishments. We had few rights, and our contracts lasted a year only — after this time only a handful of us would be made permanent.¹ Evidently, the managers exploited this, to try to get us to work harder, without complaining too much, and don’t even mention unionising. But after some weeks in the job, fully aware what a shit job we had, our biggest fear wasn’t unemployment at the end of the contract, but not meeting the three minutes mark and being banished to a section called “Billing Complaints”.

When you were sent to this gulag your task was to investigate suspicious bills, and call the customers back to convince them that our system was infalible, that if we had billed them for a certain number they had made the call.² In all fairness, this was often the case, and many clients would simply forget having called a relative, a friend, some business — it’d often just take telling them the name of whoever owned the number for them to remember and recant. But the system was far from flawless, and I’ve seen bills so high that no human being would have been able to call so many numbers, so fast, so consistently. Trying to convince someone that they weren’t being robbed was a depressing fate, but at least it wasn’t as disempowering as hearing everyone repeat the same, all day long: “It was better when I could talk to someone in person.” This was disempowering because they were right, and acknowledging our uselessness made our jobs even less fulfilling. Obviously, the process of depersonalisation the company was going through wasn’t meant to improve customer services — it was designed to make it easier to rip clients off, and to deliver more profit. It worked a marvel. And we were accomplices in this robbery.³

Little did I imagine then that these faceless arrangements would eventually become the norm.

Working for the machine, for free

I sometimes feel that a large part of my life in the past few years has been spent standing in front a self-checkout at a supermarket, listening to a dumb machine tell me that “there’s an unexpected item in the bagging area”. I try to avoid these Molochs, but sometimes I have no other choice than surrendering to them, either because the queue to see a human being is too long, or because there aren’t humans in sight. My dislike for self-checkouts goes beyond the lack of human contact — it’s also about the free labour they entail. The perversity of having someone both pay for their shopping and work for the supermarket, as a de facto cashier, is outrageous.

But this is the stage of faceless capitalism we are at. The system, no longer happy with just hijacking your money or labour, now makes you work in the process of spending your money. Self-checkouts, questionnaires, loyalty programmes, online reviews and ratings, product registrations, product assembly and set up — there’s now a wide array of ways in which we become unpaid labourers in the process of spending our money or generating profit for someone else.⁴ Capitalists might argue that by transferring this labour to the customer the product becomes cheaper, but this is hardly ever the case. Do you pay less for a product when you use a self-checkout?

You may. If you steal your shopping. I won’t be the one to condemn you for your nimble fingers.

The Digital Labyrinth

Some weeks ago I found myself trapped in a labyrinth. This labyrinth lacked hedges and gravel — it was made of ones and zeros and sound. Also known as “interactive voice response systems” (IVRs), these digital labyrinths now form part of our daily life. “If you are calling for X, press one. If you are calling for Y, press two.” You know the drill. Navigating these entanglements often implies keying in customer or account numbers, responding to security questions, providing letters from memorable words, etc. All which will be requested again, once you’ve managed to get to a human being. If you don’t give up and leave at the labyrinth’s entrance, that is, you hang up.

I can’t remember now which specific labyrinth I was trapped in — it could have been my bank or an insurance company — but in this particular one I wasn’t supposed to press buttons. Instead I was asked to speak to a faceless Sphinx. “What is the reason for your call?”, was the opening question. Unfortunately, Sphinx had trouble understanding my accent. I was asked the same question several times, always followed by “I’m sorry, I don’t understand. Can you say that again?”. At the end I ended up screaming “I WANT TO SPEAK TO A HUMAN BEING”, over and over, until I was put to a human being, who asked me to repeat my reason for calling, my account numbers, the letters from my memorable word.

It is rather obvious that the purpose of IVRs isn’t so much to direct callers to the right place but — much like a conventional labyrinth — to send a person into a complex, confusing, and time-consuming journey. Whoever is strong enough to reach the end of the digital labyrinth — which is only the beginning of the expedition inside the customer service complex — will be so weakened and irritated by the automated nonsense that will feel a sense of accomplishment just by talking to a fellow human. You approach the beginning of the battle already tired and without many expectations.

What does it mean to remove the ludic aspect of a labyrinth, in order to appropriate the structure for these systems of people management? It means surrendering what was an uncommon game to the production of profit. This is a game in which you participate unwillingly. This is a perverse game.

Borges said that “there’s no need to build a labyrinth when the whole world is one”.⁵ Perhaps now he would say that “there’s no way to build a new world when we are all trapped in the Labyrinth”.

Kafkian algorithms

In Capitalist Realism: Is There No Alternative? Mark Fisher discusses Franz Kafka's novel Der Process (aka The Trial or The Process) as a metaphor for the experience of living under “capitalist realism”.⁶ Fisher unpacks the bureaucratic absurdity underpinning Kafka’s work, comparing it to the managerial society of the mid to late 2000s — an absurdity ironically resulting from the ideology of efficiency embedded in late capitalist society. I say “ironically”, because there’s no better way to waste one’s time and to be less efficient than getting lost in bureaucracy.⁷

Admin has long ago taken over our lives, and there’s no number of specialists you could hire to get away from it. Say you have an accountant to help you complete your tax return, then there’ll be admin you’ll have to fulfil, so that the accountant can finish your tax return. And so on. Contemporary life is overwhelmed by meaningless tasks — everything now requires several forms. Needless to say, that every job is now saturated with admin, even manual jobs. Admin is everywhere, everywhere there’s admin. Contemporary life feels like one of those Matryoshka dolls. But instead of dolls within dolls what you find in our Matryoshkas is admin within admin.

What makes our moment of “capitalist realism” much worse than Fisher’s times is a development that we could call the normalisation of algorithmic culture. More and more algorithms mediate our lives, mainly thanks to the virtual playgrounds of social media. But the normalisation doesn’t end there. It is common in 2024 for algorithms to be used as tools of governance, with governments and corporations increasingly relying on algorithmic systems, in order to make decisions, from the most banal to the most important. For example, in 2023 it was revealed that the British Department for Work and Pensions (DWP) had committed £70m in technology to tackle benefit fraud, set to be invested from 2022 to 2025. This technological zeal follows accusations that the DWP is unfairly targeting disabled people for alleged “benefit fraud”⁸ — as always, these systems are far from infallible. But as AI continues to improve its capabilities, it is justified to imagine the trend to automatise vital processes will grow even more.

Will all this faceless algorithmic decision-making eventually improve our lives, or simply continue to reinforce the dominant ideology of our times — an ideology in which efficiency and productivity are prioritised over considerations of social justice, equality, and well-being?

I feel there are few reasons to be optimistic.⁹

---

The Leftovers is a reader-supported project. Keep me writing with a premium subscription or hitting the button below.

1. This type of contract — called “pasantía” (apprenticeship) in Argentina — was introduced by the neoliberal administration that governed the country from 1989 to 1999, as a way to flexibilise work. In many ways “pasantías” anticipated the zero-hour contracts that are ubiquitous today all over the world.

2. This blind trust in IT systems is an epochal trait of the 1990s and early 2000s. See the Post Office and the Horizon IT system scandal.

3. We didn’t even use our real names. I was “Dante” in this job. I’m not sure in which circle of Hell I’d place the call centre, but the name was very appropriate.

4. The most pathetic example being users paying for Twitter verification, so that they can produce content for free.

5. “No hay necesidad de construir un laberinto cuando todo el universo es uno”. Attributed to Borges. Like many Borges quotes it might be fake.

6. To summarise, a form of capitalism that presents itself as the only alternative.

7. Fisher explains very well how the admin supposed to make production more efficient becomes the end in itself, actually impairing productivity. See chapter 6, “All that is solid melts into PR: Market Stalinism and bureaucratic anti-production”.

8. “The DWP algorithm seems to send poor and disabled people to the call centre from hell,” said a representative of campaign group Foxglove. Rings a bell?

9. I can imagine Kafka’s Der Process adapted to our age. An algorithm picks Josef K’s name according to obscure and pre-determined criteria. Josef then spends the rest of the book trying to get hold of anonymous legal clerks, failing to navigate an interactive voice response system. Josef will not only remain ignorant of the reasons for his process, but now he’ll be punished impersonally: by getting his bank account frozen, or having his benefits cut. At the end of the novel, when Josef gives up all attempt to fight back the faceless monster, instead of getting stabbed by executioners, he ends up alone in a remote non-place, repeating the mantra “It was better when I could talk to someone in person… It was better when I could talk to someone in person… It was better when I could talk to someone in person…” In our age, the executioner has been replaced by the self-checkout, if you know what I mean.