…and why you should be happy about it.
I think once every two years is a good cadence for publishing blog posts, don’t you?
This post has been prompted by a couple of things. The first is the book I’ve just finished reading (more on that in a bit). The second is an issue raging in the UK at the moment about the pressure on the National Health Service (NHS) and, more specifically, the difficulties faced by everyday doctors, or General Practitioners (GPs) as they’re commonly known. That issue involves GPs being put under pressure to work longer hours, telling sick patients not to come to the surgery, and the potential of the UK hiring hundreds of GPs from overseas.
Most people with experience of it would probably agree that a trip to the doctor in the UK can be a painful experience (whatever your symptoms). Though that’s in no way to denigrate the skills and dedication of the vast majority of GPs. It’s also not to criticize them to suggest that they could, and perhaps should, be replaced by bots (and here’s a good piece to get yourself across the reasons why bots will be huge over the coming months and years). Put simply, a trip to the doctor is an algorithmic process, and can therefore be solved by, well, an algorithm. A patient presents a set of symptoms, these are examined against the GPs education and experience, a diagnosis is made and a suitable treatment prescribed. There’s nothing in there that a decent bit of artificial intelligence (AI) linked to a load of data couldn’t do more quickly and accurately than a human being.
Sceptical? Here are a couple of scenarios. The first one I’m sure many can relate to. The second a view of the brave new bot-driven world.
Both scenarios start the same way: you wake up on a Monday morning feeling pretty dreadful. Sore throat, fever, aches…certainly in no shape to get into work, and definitely in need of decent proper medical advice and treatment.
In scenario one, you have a choice to make. Do you start ringing your local surgery trying to book an appointment, or do you drag yourself down there and hope to get a walk-in? Either way, if you’re lucky, a while later you find yourself sitting for longer than expected in a waiting room packed with other sick people until you’re invited to see the doctor. You’re in front of the doctor for less than 10 minutes. She’s stressed, possibly feeling under the weather herself, and has a short time to listen to your symptoms, glance over your medical record, make a diagnosis, and prescribe you a treatment that will hopefully suit your physiological make-up. A couple of hours after you left home (again, if you’re lucky), following a trip to the pharmacy, you’re back at home, almost certainly feeling worse than when you left.
In scenario two, you grab your phone, fire up the NHS chatbot, log in, and start inputting your symptoms. The bot asks increasingly specific questions based on your responses (and has all the time in the world to do so), compares your symptoms against those of millions of others worldwide, and narrows down to an accurate diagnosis. Then, based on your entire health record, your sequenced genomic profile, perhaps the genomic records or your parents (which you’ve granted it permission to access), and millions of similar genomic profiles, it prescribes a course of treatment. It sends the prescription to your mobile phone for you to nip out and pick up at the local automated pharmacy, or perhaps even get delivered directly to your door. Probably by a drone.
This scenario is obviously quicker, more comfortable for you, and almost certainly results in a more accurate diagnosis and, therefore, effective treatment.
This is something that should happen within the next few years. In fact, for a student starting medical school this year – and maybe even one who’s a couple of years in – the prospects on graduation might be pretty bleak. Or at least, in my view, should point to taking a slightly different direction.
The scenario above is focused on the overloaded GP: the first port of call for anyone who’s feeling unwell, at least if it’s not serious enough to warrant a trip to accident & emergency (A&E). And I reckon the bot solution would probably address the symptoms and sickness of about 80% of people that visit their GP on a daily basis. This obviously takes a huge burden off GPs, but intelligent bots should also result in more people being steered towards a specialist medical practitioner more quickly. So if I was a medical student today, I’d be thinking about what I’m going to specialise in, rather than becoming the generalist.
Now, back to that book. This post, in broad terms, reflects the theme explored in a startlingly thought-provoking way in Homo Deus, by Yuval Noah Harari. In short, Homo Deus looks at where our current technological trajectory might take humankind. But it’s not solely forward-looking. In fact, most of the book deals with the historic references that point to how humans deal with advances in society, science and technology. It’s an amazing book. Advances in technology have a history of replacing the utility and value of animals (the internal combustion engine and horses is an obvious example), and there’s a lot of evidence that technology is starting to replace the value and utility of humans. You won’t like some of what’s in it, you almost certainly won’t agree with everything, but it’s a brilliantly constructed argument, and beautifully written. Recommended.