AI: A Pessimistic Take in three parts. Based on deep-seated, unarticulated fears and observations about professional erosion due to AI.

The doom scenarios of AI often remind me of those worst-case scenario maps that we are sometimes presented with to illustrate what the world would look like if all the ice caps in the world melted. You know the maps I mean: the kind of animated projection maps, like NOAA’s Sea Level Rise Viewer, showing what the map of the world could look like in 2070, 2100, 2200, and so on. In other words, what a rise of up to 70 metres would mean over time. You see a third of China’s coastline gone, so is southern England, most of north-western Europe, Bangladesh, the Niger Delta, Manhattan, San Francisco, etc., all under water. Siberia hugely shrunken. Some places are consumed from within: Australia and Florida.

Florida…all that’s left is just a shattered archipelago, a bit like the Keys, except that these are also underwater. Where there is land, there will be sea, and the ridges and hilltops of now will form future islands. Florida is already losing ground to the rising sea. Its limestone foundation lets the ocean in from beneath.

Let’s consider the now. Imagine the current landmass of Florida represents jobs that humans do. The water is AI. And just as these worst-case climate change scenarios may not happen overnight, neither will AI crash in like a sudden catastrophe, similar to how tsunamis are portrayed in movies. It rises slowly. Quiet and creeping. It seeps first into the edges – customer service, data entry, diagnostics – and then begins to swallow the middle: jobs built on terrain similar to the limestone of Florida: porous and vulnerable. Jobs involving repetition, process and even specific expertise, slowly flattened by automation. The robots do not attack. They seep in, they absorb, they replace.

By this time, the erosion of whole professions is underway. A litany, a laundry list that could mimic – to my ears at least - the repetitive, staccato rhythm of the last verses of The Proclaimer’s ‘Letter From America’:

“Accountants no more,

Paralegals no more,

Cashiers no more,

Translators no more…”

You get the idea. This is already happening. Job losses through automation are not a threat; we are several decades into it. We did not realise because the AI revolution was never shared. But it was definitely very well signposted, and we all fed the Beast. For instance, consider the many translators who took shortcuts – when they contractually should not have – and used Google Translate. Millions upon millions of people, for two and a half decades, feeding in trillions of words – words that, by the way, were always proprietary, but they ignored that minor inconvenience because, well, who was ever going to know? The same is happening with tools like ChatGPT.

And now there is a human cost. Where there was land – jobs that humans did – there are only rising seas. And we can only run on for a long time, just like the old Gospel song says and we can scramble onto higher ground and ask the rocks to hide us. They won’t. What is left is a scattered archipelago of outcrops: jobs that still need humans. Jobs that require creativity, dealing with ambiguity and emotional intelligence. Artists, educators, strategists, therapists and, of course, Patient Voice researchers.

The trouble is, now we have service refugees: people who have already seen their original roles erode and disappear due to AI. They are looking for a new ‘something’. In the sectors I am in, this means moving into services like linguistic validation, readability testing and Patient Voice research. These displaced workers pivot into these spaces, armed with short courses, repackaged bios (embellished CVs), and new narratives (possible misrepresentation of self) and AI-generated LinkedIn posts.

Some of them are welcome additions, bringing in a different angle, a different skill set that benefits the discipline as a whole. But, equally, the rest is a mixed bag, mostly of imposters and quislings. What unites all of them is a distinct lack of understanding about any foundations of what patient research is.

These islands were not supposed to hold this many people. They were built for depth, volume and space to think and work in. Not for elbows-out jostling. This is not the place for a hustle. And the water is still rising and more of the islands are being submerged, as AI consumes more human activity. For example, imagine if AI starts doing the entire analysis of qualitative data or bots being used as interviewers – depressingly, eliciting data much better than a human. In this scenario, a Patient Voice research team reduces from 40 plus to less than 5 people and this is repeated all across the sector. The tide is very high now. And, still, the service refugees flood in.

In this future scenario, procurement teams might not be able to tell the imposters apart from the real talent and, by now, nobody cares. However, the imposters themselves do. They know. They sniff your talent out with the nose of a predator and, just like in the movie, There Will Be Blood, they will drink from your milkshake.

Sales tactics become even more underhanded and duplicitous. Long-tongued liars. Midnight riders. Gamblers and ramblers and backbiters (to quote the song again). It is a stampede. A swarm. A plague, stripping everything bare.

And the sectors in question are altered, forever.

This article has focused on pessimistic scenarios of professional erosion due to AI. The next article looks at the price society will pay for this erosion happening en masse.

Thank you for reading,

Mark Gibson, Leeds,

United Kingdom, Easter 2025

References:

Run On For A Long Time - Bill Landford And The Landfordaires

Nina Simone - Sinnerman