AI replacing people: the question is not if but who

While waiting for my WhatsApp-based support person to respond to my request for re-attaching my security cameras to the new internet in my home, it occurred to me that it might not be a person at all, but an AI chatbot. The responses were polished, upbeat but neutral, fully informed and concise. I was thinking it was probably either a very experienced or well-trained Support rep…or a well-trained (partly through ‘experience’) AI. The quickly following question was: did I care?

This provoked an interesting thread of thoughts.

In terms of getting the tedious job of migrating my cameras to a new Wifi done I didn’t care. But then how much would, or should, another person care about my cameras and Wifi? I have worked in Customer Support, in both the IT helpdesk sense and the “would you like fries with that?” sense, so whenever I have a mildly shitty customer service experience while trying to get something tedious done an empathetic part of me says “well, if I were doing their job all day I’d probably feel pretty unrewarded and misunderstood”.

After all, someone expert enough to provide me with a seamless customer experience in accomplishing something so mundane is probably not being paid enough. They’re being paid the same as the fuckwits who give just enough of a shit to keep their job. They just give more of a shit, either in the vain hope of job progression or to alleviate the boredom. Until they don’t. Then their behaviour goes as crappy as the rest. Whereas an AI…

An AI doesn’t sleep. It learns at the same relentless pace. It is trained to give a shit and the good ones are tuned to be just enthusiastic enough without being sycophantic. If you’re a Brit. If you’re a Yank, dial up the sycophancy a bit, as theytend to love or at least expect it. A thousand slightly different personalities, but with just the right traits to do the tedious job well. Also, this training is finely tuned and consistently scalable. Not just scalable like an Indian callcentre, but consistent: these AI personas don’t have shitty mornings, or family issues. They come to work in a perfectly tuned and refined frame of mind.

So, what about all those callcentre people?

Let’s break it down. Why are they there? Because they always had ambition to talk for hours about people’s security camera Wifi issues? Of course not. It’s for the money, in exchange for doing a tedious job.

I think many professionals, particularly in the tech and AI space, tend to assume that people do menial jobs either because they’re too dumb for smart jobs or because they’re too lazy - they didn’t study hard enough. The reality, as some coders in the tech space are learning, is that sometimes there are too many people for the ‘right’ jobs. The world only needs so many (a diminishing number, seemingly) coders… or musicians, or actors. In my youth I was an actor in the National Youth Theatre. I pondered going pro, but my realisation was that there’s no shortage of acting talent. There are 10,000 Tom Cruises out there, but there’s only room for one Tom Cruise. The slope to a successful acting career is deeply inversely parabolic and littered with very talented failures. When I mentioned to some of my acting peers that I was heading to university to study software engineering, some remarked “I wish I had the brains for that,” while others scoffed at the perceived dull life pursuit. Some of them went on to be stars or high end professionals in the performing arts, some could still be waiting tables, for all I know. My point is that nobody wants most of the shitty jobs that AI could replace, but they still mostly likely need them.

Here are some fun stats:

• One in four workers globally are in occupations with some exposure to generative AI. But only 3.3% of global employment falls into the highest-exposure category. ilo

• McKinsey estimates that up to 50% of customer-care activity could be automated, potentially increasing productivity by 30–45% of current function costs.mckinsey

• The World Economic Forum projects that, by 2030, labour-market transformation could create 170 million jobs and displace 92 million, producing a net increase of 78 million. But with 22% of today’s formal jobs disrupted.

• The ILO identifies clerical work as the most exposed occupational group, while UK government analysis specifically lists call-centre, customer-service, telephone-sales and administrative occupations among those with high automation potential.publishing

The question, then, is not simply whether AI can do the job. It is which parts of the job remain, who gets to do them, and whether there will still be enough human roles left for people who currently depend on routine support work.

AI is often better suited to predictable, low-stakes, repetitive work, while humans remain important when judgement, accountability or genuine empathy matter. Perhaps we can finally automate all those shitty, menial roles and actually shift and compensate people into more human roles, like teaching, the arts and healthcare.