Why AI can make work painful: see customer service

Laetitia@Work #71

Hi everyone,

When it comes to AI and work, there are 3 fundamental questions:

#1 Will AI make me redundant? That’s usually the question most tackled. The argument often put forward is that AI complements humans more often than it replaces them. In truth complementarity and substitution are happening at once, depending on jobs and sectors, short term or medium term.

#2 Will AI make my work less valuable? (i.e. will I lose my livelihood?) If my job can now be performed by somebody with less skills then I’ll have to compete with people who are paid less. Will more of tomorrow’s jobs be low-paid? This question is as important as the number of jobs in the future. (Not all jobs are equal).

#3 Will AI make my work more painful? Automation is usually assumed to make work easier as machines are expected to take on boring, repetitive or physically hard tasks. But AI doesn’t always make work easier and more comfortable. What if AI takes on the tasks with which you can learn a trade? Or the most creative and fun tasks? Or the easy ones that allow you to rest once in a while?

It is this third question that I’d like to address in this newsletter using the example of customer service and support services whose use of AI has enraged users and made work harder for the humans who remain employed in those services.

As users, many of you may have gotten frustrated with automated voice mail systems and wondered how on earth to speak to a human to resolve your unique issue. You’re not alone. In fact, the enshittification of customer service is fascinating (and terrifying) in more than one way. 

It’s not just the users who feel the loss of something. The workers still involved in customer service have also seen their work enshittified. The pace has intensified. The "easy" tasks have been automated. The remaining tasks involve angry customers and difficult cases (the ones that are not easily automated).💡👇

The enshittification of customer and support services

How did customer service get so bad? First came the waves of outsourcing, with call centres in cheaper countries. Customer service was identified as a source of costs. Indeed it’s hard to increase productivity in these services: increasing the pace will often lower the quality of the service. So outsourcing these services in poorer countries was an obvious solution. Then came automated phone triage systems (“if you are calling for this, press 1 ; if you are calling for that, press 2; etc”) and longer and longer waiting times to encourage more customers to give up and handle the problem by themselves, i.e. outsourcing the customer service to the customers themselves, better equipped to handle their own problems with Q/A pages and Youtube tutorials. 

Then came AI-powered chatbots. Every customer service department jumped on that solution with the belief that it would save more money and perhaps even improve service. The bots can answer all the easy questions and even chat with customers who have time to waste. What a boon! Even in poorer countries, human-powered services will remain fairly expensive to run. And people are harder and harder to recruit. So let’s just automate the shit out of customer service! 

It may sound cynical, but staffing call centres with enough real people to answer phones quickly, and to take the time to provide thorough and empathic support, is costly. It’s simply cheaper for businesses to have poor customer service. This includes the pesky AI chatbots you might have come across – which are much cheaper than human labour.

To get even more cynical, American researchers argue that, for US businesses with terrible customer service, the decision to leave phones ringing into the abyss is a strategy to minimise the amount of redress they have to pay consumers. The researchers suggest these businesses are “forcing customers to jump through hoops” to deter them from complaining – and claiming compensation.

It seems some businesses may not care much about their reputation either. For instance, when a business has a big enough market share, or is one of a few companies that controls an essential service, customer retention isn’t a high priority. (The Guardian)

Customers' time does not cost corporations money. The same is happening or has already happened in support services within organisations. After all, employees are the “customers” of such services and these are also costly to run. How much have IT or HR services been enshittified? It’s unclear. But it seems this enshittification process is indeed happening. More and more employees complain about lacking IT support or the absence of human contact with their HR department. Of course some tasks can be automated in support, but the disappearance of human connections and human emotional support has hidden costs in terms of well-being and productivity. Some employees had better not have an issue with their computer or their pay slip because it will cost them time, patience and energy to get the support they need.

Rage, emotional drain & cognitive overload: AI can make work more exhausting

Users (customers and employees) are enraged by the enshittification of customer and support services. This has an impact on workers too. The users of support services are not finding the support they need to be productive. And the human workers who run these services (because many people still work in those services) are left to deal with enraged customers and employees, which is emotionally draining.

Rage stems from a combination of factors: dehumanisation, extreme individualism, being trapped in one's own bubble, lack of agency, and the dehumanisation of others. It’s possible the increasing use of AI in the workplace is exacerbating these issues, potentially leading to more rage among employees and clients alike. When people call automated services, they are frequently subjected to long waits with annoying hold music and a lack of personal interaction. Mostly the experience makes them feel dehumanised and insignificant. When they do get a hold of a human employee (after a long wait), some will lash out on them and unleash all the frustration they’ve accumulated. 

Like nurses and flight attendants who have to deal with angrier, more disrespectful people, customer and support service employees find themselves with emotionally draining work. They too spend more time feeling frustrated, powerless, disrespected, drained. And it’s not just enraged customers they have to deal with! It’s also more complicated problems as only the most complex and specific cases are escalated to human representatives, leaving the rest to be managed by AI systems. Instead of balancing out easy and difficult problems, they’re left with only the hard things.

Generally, this is a significant issue: in the collection of tasks that make up a job (and a workday), there are demanding tasks that require concentration and more "superficial" tasks that can be done when you’re tired. Some tasks allow you to learn, while others let you demonstrate your expertise. So automating routine tasks, as appealing as it may seem, removes the more restful tasks. Consistently performing only difficult tasks results in a significant increase in work pace and difficulty. It’s unsustainable to engage in "deep work" or emotional labor for the entirety of one's workday!

Automation promises to handle repetitive and mundane tasks, theoretically freeing up time for more complex and value-added activities but the elimination of routine tasks means that workers are left with only the most challenging aspects of their jobs. These demanding tasks require sustained attention, critical thinking, and emotional resilience. The absence of easier tasks, which provide necessary mental breaks, can lead to burnout and decreased job satisfaction. Employees need a balance of tasks that vary in intensity and complexity to maintain their overall well-being and productivity.

It is crucial to maintain a mix of tasks for work to be sustainable and enjoyable, but also for people to be able to learn their trade. A well-rounded job includes tasks that help employees build their skills progressively. Learning a trade or mastering a profession is not only about tackling complex problems but also about understanding the nuances that come with repetitive tasks. These tasks often teach the foundational skills and attention to detail that are necessary for more advanced work.

More importantly, the "superficial," apparently unproductive human interactions that occur in customer and support services are actually essential. They provide significant value beyond immediate productivity metrics. These interactions are a crucial source of well-being for users and workers alike. For customers, being able to talk to a human being when they need assistance can make a substantial difference in their overall experience and satisfaction with a service. For workers, these interactions can provide a sense of purpose and accomplishment, as well as opportunities to develop soft skills such as communication, empathy, and problem-solving. 

In conclusion, this third question “will AI make my job more painful?” is usually the one that’s left out. And yet it is an important one. When implementing automation, we should all carefully consider the transformed mix of tasks: does the work still provide intellectual stimulation? Are there enough opportunities for learning and for rest? A workforce that only encounters clients or employees at their most irate and disillusioned moments can only lead to an escalation in hostility and dissatisfaction.

🚀 📣 Caroline Taconet, Katerina Zekopoulos and I have released 2 new episodes of our podcast Vieilles en puissance, at the intersection of three themes: age, money, and women (in French).

The 7th episode, with the 3 of us: Les vieilles en puissance, c'est nous : bilan à mi-parcours 🎧

The 8th episode, with Philippa Motte: Quels liens entre santé financière et santé mentale ?👇

👉 SUBSCRIBE NOW TO THE VIEILLES EN PUISSANCE NEWSLETTER!

---

💡Check out the latest articles I wrote for Welcome to the Jungle: Can companies avoid a brain drain as retirements surge? (in English), Pourquoi la solitude de vos salariés grève leur engagement (et comment y remédier) ? (in French).

🎙️ There’s one new Nouveau Départ episode: En finir avec le travail “low cost” (with Bruno Palier) 🎧 (in 🇫🇷). Subscribe to receive our future podcasts directly in your inbox!

Miscellaneous

⛓️ End Legal Slavery in the United States, Andrew Ross, Tommaso Bardelli and Aiyuba Thomas, The New York Times, June 2024, “Many people do not realize that Emancipation did not legally end slavery in the United States, however. The 13th Amendment — the culmination of centuries of resistance by enslaved people, a lifetime of abolitionist campaigning and a bloody civil war — prohibited involuntary servitude “except as a punishment for crime whereof the party shall have been duly convicted.” In the North, that so-called exception clause was interpreted as allowing the private contracting of forced prison labor”

To balance out the difficult tasks, you need the easy ones.