By Christine Swan

I had no intention of writing this piece today. In fact, I had another post in planning but events have rather taken over. I want to talk to you about customer service helplines.

I am not a Luddite, no matter how much I appear to be. As my day job, I am a lecturer of Computer Science and IT Secondary Education. I am a specialist in teaching trainee teachers how to make the very best use of Artificial Intelligence (AI), to help them to plan, assess and assist the pupils that they teach. I also have some expertise in technical architecture and training of Large Language Models (LLMs). So why is it that I find some uses in the wider world, so infuriatingly naff?

I had placed an online order that qualified for free delivery, and received an email estimating the delivery date and helpfully informing me that they, the delivery company, were expecting my parcel, that they had received my parcel, and that it was then out for delivery. So far, so good. Another email informed me that there would be a delay.

Ever the optimist, I assumed that this would be a teeny-tiny delay, perhaps an hour or two, maybe a day at most. Today, I was informed that my parcel had become “damaged beyond repair” – quelle horreur! But wait, how can specialist cat food become damaged beyond repair? My mind began to create visions of cans of cat food being run over, crushed by an industrial press, the driver being torn to shreds by marauding tom cats.

“We need you to contact the seller” was the advice from the delivery company, although I was still curious as to the fate of the damaged tins. The delivery company has a helpline. The assistant had a name – Ezra. They prompted me to enter the twenty digit tracking reference, from memory, as there did not appear to be any way to copy and paste the code into the chat on my phone. I copied it a few digits at a time. “Are you still there?” Ezra asked impatiently. I continued doggedly copying digits and then verifying that the sequence was correct. Finally! Ezra responded that the parcel had been damaged beyond repair and returned to the warehouse. I was also advised to contact the seller. Was there anything else that Ezra could do to help? I couldn’t answer that question because the only options were yes or no and I had some different words in mind.

ITV’s Metal Mickey, cute but useless. By Johnny Edward – Metal Mickey, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=20536648

I then went to the seller’s website and entered the same details but this time, with an order number, also, non-copiable, but with fewer digits, so easier to transcribe. The nameless chatbot asked me what did I want to do, order again, or have a refund? Did they think that my hobby was ordering cat food for the fun of it, and then requesting a refund? Of course I still need the food, but, much as the late Whitney Houston wanted to dance with somebody, I needed to talk to somebody – anybody, as long as they were human. Even if they were my worst enemy, they would suffice.

Space Odyssey’s HAL 9000 – sinister and with its own agenda – By Tom Cowap – Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=103068276

“I’m sorry but I cannot read free text” was the response. I didn’t need it to read free text, I needed it to pick out a keyword and respond accordingly. I tried again: “I want to speak to a human”. That worked and the nameless bot became a spinning circle while another respondent, let’s call her Catherine, took over the Graphical User Interface and wrote a long response informing me that my parcel had been destroyed, that no photograph was provided, so it was “lost”. I can only assume that some cat owner did very well today, except it wasn’t me.

Catherine has rescheduled the delivery for tomorrow, with a different delivery company, as the first was so incompetent, the latter being my personal observation. We shall see.

My ire was amplified by not being able to communicate with a human. Customer phone helplines, based in call centres, have been reduced in favour of AI chatbots and automated phone systems. Let’s be honest, humans have been replaced by software, just as doomsters predicted. While we were predicted to be lazing around with our new found freedoms from the yoke of work, in reality, technology has a way of attempting to replace the personal touch and depersonalise interaction. I just want to talk to a human.

Metropolis’ Maschinenmensch – imposing but silent – By Horst von Harbou – This file was derived from: Horst von Harbou – Metropolis set photograph 14.jpg, Public Domain, https://commons.wikimedia.org/w/index.php?curid=146391938

I am sure that operating a support phone line can, at times, be a thankless task. However, humans can express empathy, can placate, persuade, and console. Humans listen and try to understand. I have seen examples of chatbot, robotic patients to train nurses, a “teacher” was showcased only this week in the United States. I am presenting my most sceptical look as I type this. Year 9, last period on Friday afternoon would probably reduce said robot to a pile of gibbering electronic waste. A favourite video of mine features a home help robot attempting to cook a meal. It is unable to lift a frying pan, and deposits the contents on the floor. It then proceeds to slip on the greasy food adding a copious helping of comedy.

Doctor Who’s K9 – can talk, can’t fetch a ball – By Steve Collis from Melbourne, Australia – Doctor Who Experience, CC BY 2.0, https://commons.wikimedia.org/w/index.php?curid=24306760

The previous example makes a clear point. There are some activities that AI and humanoid robots are still really bad at, and that real humans are far superior. Robots can weld and paint cars twenty four hours a day with incredible precision. AI is capable of analysing huge quantities of data, far more quickly than a human would be able to. It can predict animal population changes with various climate change scenarios, it can recognise miniscule anomalies in X-rays and MRI scans, without tiring. AI has huge potential if used ethically and appropriately. AI can be woefully inadequate when communicating with humans, emulating being one.

In fact, even within this software, I am presented with the option to improve my blog using AI. No, thank you. There are certain activities that, in my opinion, should be left to us.