Charged with the task of organising an end of term lunch for my German conversation class, prior to the Easter break, I opted for a Scottish restaurant in Hope Street in Glasgow. Table for 14, or thereabouts. The restaurant asked me to confirm numbers the day prior, so on Wednesday morning I phoned, and a young female voice asked me in a pleasant Scottish accent how she may help.
“I’ve booked a table for 14 for tomorrow. You asked me to phone you today to confirm the number, and indeed it will be 14.”
“Just a moment… No. We have no table booked for 14. Let me see if I can book you in now. No, I’m afraid there isn’t a table available. Would you like to book for an alternative date?”
Something slightly halting – but only slightly – about the young lady’s delivery gave me pause. I said, “Could I possibly speak to a real person, like, say, the manager?”
“Certainly. Putting you through.”
I spoke to the manager who quickly confirmed the booking. “See you tomorrow.”
So much for artificial intelligence. I was speaking to a robot. She was very convincing. It’s a kind of Holy Grail amongst those whose business it is to perfect these algorithms; to make the robot entirely convincing, so that you can’t tell the difference. And they are nearly there. It is Alan Turing’s imitation game. If you can’t tell that you aren’t speaking to an intelligent being, then, to all intents and purposes, you are speaking to an intelligent being. It occurs to me that if you are being addressed by a robot, and you do not know it is a robot, then you are actually being scammed. But the dividing lines in our society between probity, sharp practice, and fraud are becoming increasingly blurred.
But why would any of us want to go down this AI route? One can imagine lots of reasons. We choose the route simply because it is there, or potentially there. It’s a challenge, a refinement of technology. One might ascribe to this particular manifestation of Artificial Intelligence the attributes commonly associated with automation; it is “efficient”, and “cost-effective”. It’s the future; if we don’t grasp it, others will, and we will be left behind. It will free up professionals in many walks of life from humdrum repetitive activity, and allow them to concentrate on their core business.
AI has already infiltrated the medical consultation in General Practice. There is an article in the current British Journal of General Practice (March 2026, Volume 76, Number 764) by Roger Neighbour, a past president of the Royal College of GPs. Two’s company, three’s a crowd: AI in the consultation. It more or less says that AI is here to stay and we’d better get used to it. Reading it, I was reminded of the state of General Practice as it was about a quarter of a century ago, when computerisation was beginning to take off. I expressed misgivings when we went from paper-light, to paperless, and my colleague said to me, “It’s coming, James. It’s unstoppable. We can’t play King Canute.” I was worried that the computer in the consulting room would become a third party disrupting the sanctity of the medical consultation. Sure enough, the visual display unit began to chip in with chirpy remarks like, “Have you thought of this… Have you thought of that… don’t forget to check…” etc. Now, with AI, we will be seeing the same effect on steroids.
Roger Neighbour asked two AI systems, ChatGPT and CoPilot, what AI’s contribution to the general practice consultation would be over the next few years. Apparently AI would offer clinical decision support, differential diagnoses, appropriate investigation recommendations, and proposed management pathways (guideline compliant). The AI role was not to replace, but to augment and enhance. AI, said Co-Pilot, would free up GPs to focus on patients rather than keyboards. None of this is new. All of these things were promised with the dawn of computerisation. We therefore have the benefit of 25 years of experience to judge whether we are being offered the prescription for a panacea, or for the nostrum of a snake oil salesman.
Well, we all know how that went. You can’t even get an appointment. Maybe a receptionist with a nice Scottish accent is saying to you, “No, you haven’t made an appointment. Let me re-schedule you.”
It seems to me that we need to ask who will benefit from the imposition of such systems, not just in medicine, but in every conceivable walk of life, education, policing, transport, social care, the law, and, God help us, warfare. The answer is the same as it was in 2001. Information Technology as it was then, and Artificial Intelligence as it is now, are technologies in search of markets. The people who will benefit are the people who are peddling their wares. The seductive attraction of the product is that users will be empowered to set up a system which will run itself. Why employ somebody, when their function can be taken over by a bot? Ultimately, the marketing of AI is all about power, and the acquisition of power by a remarkably few number of people.
AI systems work by using up vast amounts of energy to trawl vast amounts of information looking for patterns. But Roger Neighbour quoted T. S. Eliot at the head of his BJGP article: “Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?” (Choruses from the Rock.) I really don’t want anybody to scan my retina and tell me what my chances are of developing dementia. But our political masters are utterly beguiled and enthralled by the promise of AI. At long last, the problems of the NHS can be solved by turning it into a perpetual motion machine. But there is no such thing as a perpetual motion machine, and life, real life, is not remotely like that. We, Homo sapiens, at our best, don’t remotely think algorithmically. We are not robots.
Lunch in Hope Street was very good. I’m so glad I didn’t take the bot’s word for it, and book elsewhere.
jamescalumcampbell 