Over the years, including before the interwebs and mobile phones, there has been a general de-skilling of the 'professional classes'. Solicitors now use their clerks and conveyancers to do their work (sometimes without adequate supervision). A great deal of teachers' work is now carried out by teaching assistants. Many companies now deal with customer issues through an obligatory help desk - usually staffed by people who have no knowledge of the companies workings, only a stubborn insistence on following a script. You could even make an argument that many MPs rely on their office staff or party instructions too much and don't ascertain the facts properly. One of the consequences of the de-skilling is that bureaucratic rules proliferate to help the minions do the job.
Another consequence is that ordinary people often feel obliged to monitor any professionals they engage very closely, or seek help on user forums rather than tangle with a help desk.
All of which leads me to speculate that we are seeing the slow death of professionalism. We are becoming a 'do it yourself' culture because those we trusted are no longer reliable.
This has definitely been my experience and not just in respect of white collar occupations. Try getting a skilled tradesman these days to reliably do *anything*.
I fear that the professional incompetence you describe is not confined to the Law (where the consequences are mainly monetary).
This reliance on technology reminded me of some seminars I attended a couple of years ago when I did a module on the Law of the Sea at a Russell Group university, as part of a taught Masters. We were about a dozen students, of whom 9 or 10 were Chinese. The Chinese contingent sat masked and silent through every class with their laptops open in front of them. I called them the Great Wall of China. It was quite unnerving because, even in the face of direct questioning by our very engaging and affable lecturer, they would remain mute. (Meaning of course that the other British student and I would say anything at all, even if it was nonsense, to simply not leave him hanging!) When I looked over at what was on the laptop screens of the masked ones, I was astonished to see that the computers were transcribing the spoken word into written word (English), absolutely perfectly, whilst the human bricks in the Wall were playing computer games on a split screen!
I always wondered afterwards, how their essays turned out.
This is a shocking report - and a frightening scenario for the future. The point regarding reading - books, learned articles - is well made. Unwillingness or incompetence to read has deep roots. 30 or so years ago I tutored a nice girl from a top school, in A-level German because she wanted to go to Oxford (yes, she got in). I was staggered by her plain disinterest in reading even short paragraphs, never mind actual articles from german newspapers.
There's a bigger problem though: general literacy has been going down the drain while at the same time we ordinary people are being taught and exhorted to look up to and praise as 'experts' all those 'professionals' who need literacy in their professional work - like that barrister in this report.
While scientific illiteracy and innumeracy has worse effects on life generally (e.g. Covid ... nuff said), it is perhaps a small glimmer of hope to realise that certain trades and certain science-based professions still teach proper attention to detail while insisting on meticulous work.
Oh how depressing! And how true! You can see in general life how incompetence is now pretty much the norm. How people no longer seem to actually know anything - not how things work, or why they work or how to fix them if they stop. People are no longer interested in improving themselves or their surroundings and everything is falling apart. I do think it's a combination of appalling educational standards, modern technology, easy welfare/benefits and general "soft" living. We are sinking and most of us don't really care.
I think there are two overlapping phenomena in this case. It is too easy to blame information technology itself, which can be used intelligently if we choose to. One case law archive I use offers 'smart summaries' which are abstracts created by machine learning. This saves time finding relevant judgments which we can then read in detail.
Since judgements are numbered sequentially, it should have been obvious that a citation cannot be faked. Perhaps the barrister in this case assumed the judge would pay as little attention to the sources as they did themselves.
Half my family comes from the area now known as Haringey. Its local government and legal context has been political, which is to say anti-meritocratic, since the 1980s at least. Please see my essay on reparations for background: https://danielhowardjames.substack.com/p/reparations-for-slavery-and-war-who
The other phenomenon I blame is 'participation trophy' education. Recently my local council approached the Church of England primary school in the area with an offer to run an imaginative competition. Entries of essays, poems or drawings would be accepted. The head teacher refused permission, on the basis that it was meant to be a competition with first and second prizes. She would only allow the pupils to take part if every one of them got an equal prize, regardless of the effort made.
That's the point I was trying to get across. It's not exactly AI that's the problem. It's that people don't want to do their jobs properly (and therefore resort to inappropriate use of AI).
Or, that large language models were designed as plagiarism machines in order to satisfy the market for quick and easy but not necessarily accurate results. There are ‘AI’ products for students which re-write the output of other AI's to deceive plagiarism-detecting AI's.
I've written previously about how Turing's work was rooted in deception, in both hot and cold wars, and so we shouldn't be surprised that the whole premise of machine learning is the imitation game, rather than supporting and amplifying human intelligence.
Very interesting comment. My feeling is that the best they can manage is pastiche. We therefore need a proper theory of pastiche - I’ve been planning a post about it for a while.
In the 1990's, friends of mine had a night shift job in which they would read every national newspaper just as it was printed, with each reader assigned to search for stories of interest to a particular client. They literally cut and pasted the relevant newspaper articles into a book, which would be couriered direct to the client in time for their breakfast.
There has been a bait and switch to persuade us to invest emotionally and financially into extremely capital and energy intensive 'AI' start-ups with flawed large language models. They are not intelligent; they represent automated white collar labour. Like early industrial automation they are bankruptingly expensive, exponentially greater in energy use, and prone to going seriously wrong.
With the advent of the public Internet, intelligence agencies realised that there was far more information available online than from spies, especially about countries or companies where no spies had been deployed previously. Traditional intelligence gathering was at a disadvantage when a new scenario developed rapidly. This realisation engendered the discipline of 'open source intelligence', or OSINT, a concept only slightly overlapping with 'open source software'. (Open source software has enabled both the breadth of online publishing and the creation of large language models).
It was a logical next step for the military intelligence community, which created ARPANET and DARPANET in the first place, to automate the work of expensive and difficult to scale OSINT researchers, from highly skilled foreign radio listeners to cut and paste readers like my friends. There is a whole genre of technical writing on this, for example: https://archive.org/details/opensourceintelligencebooksosint
BBC Monitoring is a human OSINT unit which is funded by the Foreign Office as part of the World Service with over £100 million a year. I believe it is the open sources counterpart to GCHQ's listening into confidential and private communications.
I have no inside knowledge, but I'd be willing to bet that the forthcoming redundancies of foreign language experts at the World Service are because some genius in BBC management thinks 'AI', podcasts and social media will do the same job for less money. A major tactical mistake for the UK, in my opinion.
Large language models are built on sand, because the anyone-can-publish model of the public Internet has debased the quality of the training material, and because the software has no disincentive to lie. It's designed to aim for plausibility (passing the Turing Test), not accuracy. It only takes a moderately qualified or experienced person to know when ChatGPT is spouting rubbish.
Visual machine learning is no better. In a trial of a skin cancer detection app, the results were perfect. Then the designers realised that the software had learned that a ruler in the picture meant cancer. No one was putting rulers on healthy skin to measure cancers that weren't there.
On top of that, we have the personal and business agendas of the creators of these products which deliberately skew the results or refuse to provide answers to certain questions. Programmers aren't political neutrals.
What has always struck me about ChatGPT is that produces shockingly good mediocre results. It is genuinely impressive that it can come out with what it comes out with. But what it comes out with is not impressive in itself. People get these two things confused.
A huge amount of money has been invested in creating software which can respond plausibly to natural language queries (thereby passing the Turing Test) since the failure of Ask Jeeves. It's the holy grail of Internet search, because it offers the promise of eliminating expensive and potentially unreliable white collar labour.
The idea is that the non-expert leader or executive simply asks the computer for the answer they want, rather than relying on a middle class of researchers and academics to write long reports that the important people don't have time to read. Eventually we won't need a legal profession, because the computer will know who is guilty and what is lawful.
Venture capital seeks opportunities to shrink workforces, because since the 1970s when improvement in working class lives stalled, getting rid of people with demands has been seen as the route to profitability. What can be automated is, and what cannot be automated cheaply enough is offshored to regimes which can set conditions for wage labour, either because of central control, or because of competition caused by poverty. A relative of mine with a manufacturing business has moved production from China to Vietnam because Chinese expectations rose.
But the machine no more understands the question than an impressively capable parrot understands the phrases that we teach it. Most of the current products are very American in character, because they assume that throwing more money and power at the problem of ruling class dependence on experts will provide a solution. There was a reckoning for the USA's uncompetitive automotive industry, and I predict the 'AI' start-ups will go the same way.
Conscientious people are known to have much lower birth rates as fecklessness as a trait has been selected for in modern British society for many years now. The conscientious are a dying breed everywhere. Look at Japan
‘And that’s leaving to one side the pernicious effects of other modern workplace technology - email being the chief culprit here - which constantly distract attention and cultivate a tendency to perpetually dwell in an inbox rather than develop headspace to complete tasks properly.’
Why leave this to one side? This is the key to the dumbing down and immiseration of professional life. No time to think, read, research, read cases, prepare a cogent response. No idea what can be done to fix this, maybe identifying the problem is at least a start.
Cal Newport wrote an interesting book about this called 'A World Without Email' where he shows ways to reduce email. Unlikely, as most companies won't want to change upstream systems because 'this is the way we do things here'.
Yes, I am a big fan of Cal Newport, and I’ve tried very hard to avoid checking email before noon every day partly as a crude way of implementing his ideas.
Sometimes it really is death - you remind me of this classic article about the collapse in competence at Boeing: https://www.nytimes.com/2019/09/18/magazine/boeing-737-max-crashes.html but it applies in so many different fields (including the church). I tend to the view that if you select for anything other than excellence then excellence will become increasingly scarce.
Yes indeed. First noticed it when my technically excellent small ISP was taken over by successively larger companies… they fired all the techies, lied about their service levels, put the prices up, employed foreign call centres and become absolutely useless. Now all ISPs are like that. Rinse and repeat with any (service?) industry.
Not all ISPs. If you want proper technical people on the end of the phone, try Andrews & Arnold ( I use them for my VOIP service). Take a look at their website, and you'll soon see what I mean. I called them to ask some questions about about support for the actual physical cable service (i.e. OpenReach vs City Fibre) , and got to speak to someone both knowledgable and helpful. I hope they survive.
I recently spent £350 (plus VAT) on a consultation with a provincial solicitor to confirm my understanding of the law and procedures derived from internet research prior to defending myself at a trial.
Perfectly expressed: "How can it be that people are completing fourteen years of schooling, three years of undergraduate study, another year or two of training to become a barrister or solicitor, and coming out at the end being incapable of reading or writing properly, and without any sense that tasks have to be done properly?"
Yup. This is one of the reasons I got out of the teaching racket.
…making the professional lives of teachers increasingly difficult is no doubt one of the root causes - and adding VAT to fees of paying schools is another step in the wrong direction.
Regarding research statements and the like, I sympathize with those who resort to LLMs to write them, because I myself hate writing application essays, cover letters, and other things where I have to pretend to care about things I don't really care about, or to have the kind of personality I don't really have, to get things I want for reasons that would disqualify me from getting them if those reasons were spoken aloud. If I don't really want a job because I'm interested in the work and not just because I want to avoid the disgrace of unemployment and homelessness while staying out of other people's ways as much as I can, maybe that means I don't deserve to make a living, but, not having the resolve to decide whether I should live or perish on my own, I'd like if it our civilization's leaders could make it clear to us who have qualms about falsifying our personalities and interests if we have a place in the world or if we should die.
Ha, yes, well, I think that might hold true for job applications. But if you want to do a PhD you should at least be interested in the subject enough to write your own research proposal, no?
Indeed, if you can't write the proposal how will you manage a whole thesis!
My definition of a PhD: a book length answer to an exam question you set yourself, in order to prove you are the worlds foremost expert on something nobody else can be bothered about.
That said I do have some sympathy with those faced with producing reams of bumf for research proposals… 'You must include a data management plan'; 'we'll put all the papers n' stuff in a shoe box and keep in the attic' might be a bit flippant, but filling pages with details nobody will read… the temptation to ask a robot or a trained monkey to do it is tempting…
I was talking to a young woman who because of her parents was fluent in both English and an unusual language for a British citizen. A research job regarding her parents' country of origin was advertised which she really wanted, but a masters degree was the minimum requirement. She only had one degree. I advised her to go for it anyway, as the vast majority of British applicants would be relying on Google Translate.
When in school, I was taught that the purpose of a PhD was to make an original contribution to your chosen field. I don't see how that's possible using ChatGPT. But the pipeline of new academics must be filled if the university sector is to grow.
There is a bit of a civil war emerging. There is a large contingent of academics who are embracing AI for various reasons which I think are nuts, but which I suppose to them are plausible. Having seen the bilge that LLMs can come out with I wouldn’t trust it to produce anything I wanted to put my name to.
Haringey Law Centre is not a normal firm of solicitors. It is a charity that provides legal advice. It proudly says on its website:
"As well as providing specialist legal advice, the Law Centre is active in our local community. We also work with sister Law Centres around the country to campaign on Access to Justice issues."
"Since being called to the Bar in 2021, she has effectively applied her legal expertise across various capacities that merge legislative compliance with active social intervention."
Make of all that what you will. I do think there is still substantial excellence in the legal profession, of which the judge's extremely witty incineration of Ms Forey is a great example!
No doubt there is - but it is diminishing, and tends to be found in older generations. A lot of the Tribunal decisions I read evidence a disturbing shoddiness across the piece.
Technology is not only an addiction, it is a way of life. The youngest generations feel they have to be on their phones to socialise and to keep up with the culture of their peers. I think a child who gets a phone later is one who will be confined to a much smaller group of friends.
It would also be a shame to miss out on content like this article, or some of the most educational YouTube videos through a dogmatic rejection of technology. But if I'm not going to reject technology completely, I have to contest with the diversity of content and its diversity of value. How do I know the content I am consuming was created by someone who has read first or second sources? Or someone who has condensed the information in a useful way? Or someone who has the professional integrity to write and research honestly? Articles like this one obviously rank higher than something on TikTok, but often the difference isn't so clear.
We categorise and teach forms of writing in school (expository, persuasive, narrative, and descriptive), I think it is also important to teach the differences in content quality and their value. This would be similar to the way some schools are teaching how to identify fake news. Once children are explicitly educated in the value of primary sources, educated writers and long form content, I suspect they may pay more attention to them.
Yes, I've noticed this. And I'm sure in the fulness of time we will figure out ways to better use/limit the tech. In the meantime, though, a whole generation has been very badly damaged
This is a problem in US legal practice as well. A lawyer was discovered to have used AI to write a court filing. The filing contained bogus citations. When confronted about the bogus citations, the law firm claimed to have cite checked the filing. Clearly the filing was not cite checked because cite checking is the process that was used to reveal the bogus citations.
"Speaking paradoxically we may say that incompetence, having been standardized, has now become an essential part of professional excellence. We have no longer incompetent professionals, we have professionalized incompetence."
Very interesting read. However, I would hesitate to generalise too much from a case brought by Haringey Law Centre against Haringey Council. This is a poor, very diverse, London borough and I would be surprised if either party were able to attract any brilliant or ambitious lawyers of any stripe. It is, of course, shocking that this sort of farce is being conducted with tax-payer money.
Sure - and no doubt local government has always been characterised by a certain amount of laziness and sloppiness. But this is a whole other level, and is not what tends to go on in other countries (e.g. France, Japan).
Haringey is a very wealthy Borough. It includes what was said to be the most expensive street in Britain (Bishops Avenue) and upmarket neighbourhoods such as Crouch End. Even the areas with more social housing such as Tottenham are gentrified, being within a short Tube ride of the West End. Yet when I lived there, the two bottom placed high schools in the whole of the UK were both in Haringey. Its problems are entirely political, because it has been controlled by the hard left since the 1980s.
I share your pessimism of the future given the trajectory. It's a problem that isn't limited to law, unfortunately. Other authors, thinkers, professionals, and pedagogues have noted it as well. There's a very clear link between the death of the literary mind and the death of civilization. We saw something similar in the post-Roman era, when literacy and logic were at a low point. The digital world, by nature, precludes the human mind from long, complicated thought. The very same long and complicated thought that built the modern world. So I don't think it's a stretch to insinuate that the digital world will never produce another Einstein, Oppenheimer, Newton, Montesquieu, Lock, Nietzsche, or other great and world changing thinker (and no, I do not count the modern industrialists among those; I would trade 1000 Elon Musks or Steve Jobs for one more Einstein or one more Seneca).
Conversely, it's never been easier to be "above average" since the bar is now so dismally low.
I believe the key word here is "professional".
Over the years, including before the interwebs and mobile phones, there has been a general de-skilling of the 'professional classes'. Solicitors now use their clerks and conveyancers to do their work (sometimes without adequate supervision). A great deal of teachers' work is now carried out by teaching assistants. Many companies now deal with customer issues through an obligatory help desk - usually staffed by people who have no knowledge of the companies workings, only a stubborn insistence on following a script. You could even make an argument that many MPs rely on their office staff or party instructions too much and don't ascertain the facts properly. One of the consequences of the de-skilling is that bureaucratic rules proliferate to help the minions do the job.
Another consequence is that ordinary people often feel obliged to monitor any professionals they engage very closely, or seek help on user forums rather than tangle with a help desk.
All of which leads me to speculate that we are seeing the slow death of professionalism. We are becoming a 'do it yourself' culture because those we trusted are no longer reliable.
This has definitely been my experience and not just in respect of white collar occupations. Try getting a skilled tradesman these days to reliably do *anything*.
The same applies to "walk-in" medical care.
I fear that the professional incompetence you describe is not confined to the Law (where the consequences are mainly monetary).
This reliance on technology reminded me of some seminars I attended a couple of years ago when I did a module on the Law of the Sea at a Russell Group university, as part of a taught Masters. We were about a dozen students, of whom 9 or 10 were Chinese. The Chinese contingent sat masked and silent through every class with their laptops open in front of them. I called them the Great Wall of China. It was quite unnerving because, even in the face of direct questioning by our very engaging and affable lecturer, they would remain mute. (Meaning of course that the other British student and I would say anything at all, even if it was nonsense, to simply not leave him hanging!) When I looked over at what was on the laptop screens of the masked ones, I was astonished to see that the computers were transcribing the spoken word into written word (English), absolutely perfectly, whilst the human bricks in the Wall were playing computer games on a split screen!
I always wondered afterwards, how their essays turned out.
A common experience!
🤦🏼♀️
This is a shocking report - and a frightening scenario for the future. The point regarding reading - books, learned articles - is well made. Unwillingness or incompetence to read has deep roots. 30 or so years ago I tutored a nice girl from a top school, in A-level German because she wanted to go to Oxford (yes, she got in). I was staggered by her plain disinterest in reading even short paragraphs, never mind actual articles from german newspapers.
There's a bigger problem though: general literacy has been going down the drain while at the same time we ordinary people are being taught and exhorted to look up to and praise as 'experts' all those 'professionals' who need literacy in their professional work - like that barrister in this report.
While scientific illiteracy and innumeracy has worse effects on life generally (e.g. Covid ... nuff said), it is perhaps a small glimmer of hope to realise that certain trades and certain science-based professions still teach proper attention to detail while insisting on meticulous work.
I am a 62 year old medical practitioner.
I see much the same approach in many young doctors.
On a brighter note, my solicitor son and medical registrar, both in their early thirties, both read the classics with their wives every evening.
They watch no television.
Their love of literature is profound and runs counter to the educational zeitgeist.
It gives one hope for humanity, though in isolated pockets.
Very isolated, sadly.
Oh how depressing! And how true! You can see in general life how incompetence is now pretty much the norm. How people no longer seem to actually know anything - not how things work, or why they work or how to fix them if they stop. People are no longer interested in improving themselves or their surroundings and everything is falling apart. I do think it's a combination of appalling educational standards, modern technology, easy welfare/benefits and general "soft" living. We are sinking and most of us don't really care.
They will care when it gets bad enough. I ‘reassure’ myself with that.
The more complicated life becomes, the harder it is to master. Ask the average Internet user to explain how it works, and you'll get a blank look.
I think there are two overlapping phenomena in this case. It is too easy to blame information technology itself, which can be used intelligently if we choose to. One case law archive I use offers 'smart summaries' which are abstracts created by machine learning. This saves time finding relevant judgments which we can then read in detail.
Since judgements are numbered sequentially, it should have been obvious that a citation cannot be faked. Perhaps the barrister in this case assumed the judge would pay as little attention to the sources as they did themselves.
Half my family comes from the area now known as Haringey. Its local government and legal context has been political, which is to say anti-meritocratic, since the 1980s at least. Please see my essay on reparations for background: https://danielhowardjames.substack.com/p/reparations-for-slavery-and-war-who
The other phenomenon I blame is 'participation trophy' education. Recently my local council approached the Church of England primary school in the area with an offer to run an imaginative competition. Entries of essays, poems or drawings would be accepted. The head teacher refused permission, on the basis that it was meant to be a competition with first and second prizes. She would only allow the pupils to take part if every one of them got an equal prize, regardless of the effort made.
That's the point I was trying to get across. It's not exactly AI that's the problem. It's that people don't want to do their jobs properly (and therefore resort to inappropriate use of AI).
Or, that large language models were designed as plagiarism machines in order to satisfy the market for quick and easy but not necessarily accurate results. There are ‘AI’ products for students which re-write the output of other AI's to deceive plagiarism-detecting AI's.
I've written previously about how Turing's work was rooted in deception, in both hot and cold wars, and so we shouldn't be surprised that the whole premise of machine learning is the imitation game, rather than supporting and amplifying human intelligence.
Very interesting comment. My feeling is that the best they can manage is pastiche. We therefore need a proper theory of pastiche - I’ve been planning a post about it for a while.
In the 1990's, friends of mine had a night shift job in which they would read every national newspaper just as it was printed, with each reader assigned to search for stories of interest to a particular client. They literally cut and pasted the relevant newspaper articles into a book, which would be couriered direct to the client in time for their breakfast.
There has been a bait and switch to persuade us to invest emotionally and financially into extremely capital and energy intensive 'AI' start-ups with flawed large language models. They are not intelligent; they represent automated white collar labour. Like early industrial automation they are bankruptingly expensive, exponentially greater in energy use, and prone to going seriously wrong.
With the advent of the public Internet, intelligence agencies realised that there was far more information available online than from spies, especially about countries or companies where no spies had been deployed previously. Traditional intelligence gathering was at a disadvantage when a new scenario developed rapidly. This realisation engendered the discipline of 'open source intelligence', or OSINT, a concept only slightly overlapping with 'open source software'. (Open source software has enabled both the breadth of online publishing and the creation of large language models).
It was a logical next step for the military intelligence community, which created ARPANET and DARPANET in the first place, to automate the work of expensive and difficult to scale OSINT researchers, from highly skilled foreign radio listeners to cut and paste readers like my friends. There is a whole genre of technical writing on this, for example: https://archive.org/details/opensourceintelligencebooksosint
BBC Monitoring is a human OSINT unit which is funded by the Foreign Office as part of the World Service with over £100 million a year. I believe it is the open sources counterpart to GCHQ's listening into confidential and private communications.
I have no inside knowledge, but I'd be willing to bet that the forthcoming redundancies of foreign language experts at the World Service are because some genius in BBC management thinks 'AI', podcasts and social media will do the same job for less money. A major tactical mistake for the UK, in my opinion.
https://www.bbc.co.uk/news/articles/cpql1vvdn58o
Large language models are built on sand, because the anyone-can-publish model of the public Internet has debased the quality of the training material, and because the software has no disincentive to lie. It's designed to aim for plausibility (passing the Turing Test), not accuracy. It only takes a moderately qualified or experienced person to know when ChatGPT is spouting rubbish.
Visual machine learning is no better. In a trial of a skin cancer detection app, the results were perfect. Then the designers realised that the software had learned that a ruler in the picture meant cancer. No one was putting rulers on healthy skin to measure cancers that weren't there.
On top of that, we have the personal and business agendas of the creators of these products which deliberately skew the results or refuse to provide answers to certain questions. Programmers aren't political neutrals.
What has always struck me about ChatGPT is that produces shockingly good mediocre results. It is genuinely impressive that it can come out with what it comes out with. But what it comes out with is not impressive in itself. People get these two things confused.
A huge amount of money has been invested in creating software which can respond plausibly to natural language queries (thereby passing the Turing Test) since the failure of Ask Jeeves. It's the holy grail of Internet search, because it offers the promise of eliminating expensive and potentially unreliable white collar labour.
The idea is that the non-expert leader or executive simply asks the computer for the answer they want, rather than relying on a middle class of researchers and academics to write long reports that the important people don't have time to read. Eventually we won't need a legal profession, because the computer will know who is guilty and what is lawful.
Venture capital seeks opportunities to shrink workforces, because since the 1970s when improvement in working class lives stalled, getting rid of people with demands has been seen as the route to profitability. What can be automated is, and what cannot be automated cheaply enough is offshored to regimes which can set conditions for wage labour, either because of central control, or because of competition caused by poverty. A relative of mine with a manufacturing business has moved production from China to Vietnam because Chinese expectations rose.
But the machine no more understands the question than an impressively capable parrot understands the phrases that we teach it. Most of the current products are very American in character, because they assume that throwing more money and power at the problem of ruling class dependence on experts will provide a solution. There was a reckoning for the USA's uncompetitive automotive industry, and I predict the 'AI' start-ups will go the same way.
Conscientious people are known to have much lower birth rates as fecklessness as a trait has been selected for in modern British society for many years now. The conscientious are a dying breed everywhere. Look at Japan
‘And that’s leaving to one side the pernicious effects of other modern workplace technology - email being the chief culprit here - which constantly distract attention and cultivate a tendency to perpetually dwell in an inbox rather than develop headspace to complete tasks properly.’
Why leave this to one side? This is the key to the dumbing down and immiseration of professional life. No time to think, read, research, read cases, prepare a cogent response. No idea what can be done to fix this, maybe identifying the problem is at least a start.
I agree. I was only leaving it to one side to avoid the post becoming extremel long indeed!
Cal Newport wrote an interesting book about this called 'A World Without Email' where he shows ways to reduce email. Unlikely, as most companies won't want to change upstream systems because 'this is the way we do things here'.
Yes, I am a big fan of Cal Newport, and I’ve tried very hard to avoid checking email before noon every day partly as a crude way of implementing his ideas.
Sometimes it really is death - you remind me of this classic article about the collapse in competence at Boeing: https://www.nytimes.com/2019/09/18/magazine/boeing-737-max-crashes.html but it applies in so many different fields (including the church). I tend to the view that if you select for anything other than excellence then excellence will become increasingly scarce.
I came to the conclusion over twenty ago (from observations of dealing with many different organisations) that:
Mediocrity is a better business model than excellence.
Akin to Gresham’s Law. Dave’s Law !
As in mediocrity drives out excellence? (bad money drives out good?)
Yes indeed. First noticed it when my technically excellent small ISP was taken over by successively larger companies… they fired all the techies, lied about their service levels, put the prices up, employed foreign call centres and become absolutely useless. Now all ISPs are like that. Rinse and repeat with any (service?) industry.
Not all ISPs. If you want proper technical people on the end of the phone, try Andrews & Arnold ( I use them for my VOIP service). Take a look at their website, and you'll soon see what I mean. I called them to ask some questions about about support for the actual physical cable service (i.e. OpenReach vs City Fibre) , and got to speak to someone both knowledgable and helpful. I hope they survive.
I recently spent £350 (plus VAT) on a consultation with a provincial solicitor to confirm my understanding of the law and procedures derived from internet research prior to defending myself at a trial.
Perfectly expressed: "How can it be that people are completing fourteen years of schooling, three years of undergraduate study, another year or two of training to become a barrister or solicitor, and coming out at the end being incapable of reading or writing properly, and without any sense that tasks have to be done properly?"
Yup. This is one of the reasons I got out of the teaching racket.
…making the professional lives of teachers increasingly difficult is no doubt one of the root causes - and adding VAT to fees of paying schools is another step in the wrong direction.
Regarding research statements and the like, I sympathize with those who resort to LLMs to write them, because I myself hate writing application essays, cover letters, and other things where I have to pretend to care about things I don't really care about, or to have the kind of personality I don't really have, to get things I want for reasons that would disqualify me from getting them if those reasons were spoken aloud. If I don't really want a job because I'm interested in the work and not just because I want to avoid the disgrace of unemployment and homelessness while staying out of other people's ways as much as I can, maybe that means I don't deserve to make a living, but, not having the resolve to decide whether I should live or perish on my own, I'd like if it our civilization's leaders could make it clear to us who have qualms about falsifying our personalities and interests if we have a place in the world or if we should die.
Ha, yes, well, I think that might hold true for job applications. But if you want to do a PhD you should at least be interested in the subject enough to write your own research proposal, no?
Indeed, if you can't write the proposal how will you manage a whole thesis!
My definition of a PhD: a book length answer to an exam question you set yourself, in order to prove you are the worlds foremost expert on something nobody else can be bothered about.
That said I do have some sympathy with those faced with producing reams of bumf for research proposals… 'You must include a data management plan'; 'we'll put all the papers n' stuff in a shoe box and keep in the attic' might be a bit flippant, but filling pages with details nobody will read… the temptation to ask a robot or a trained monkey to do it is tempting…
Yes, it can be soul-destroying. But showing you can do a soul-destroying task well is also part of doing a PhD!
I was talking to a young woman who because of her parents was fluent in both English and an unusual language for a British citizen. A research job regarding her parents' country of origin was advertised which she really wanted, but a masters degree was the minimum requirement. She only had one degree. I advised her to go for it anyway, as the vast majority of British applicants would be relying on Google Translate.
When in school, I was taught that the purpose of a PhD was to make an original contribution to your chosen field. I don't see how that's possible using ChatGPT. But the pipeline of new academics must be filled if the university sector is to grow.
There is a bit of a civil war emerging. There is a large contingent of academics who are embracing AI for various reasons which I think are nuts, but which I suppose to them are plausible. Having seen the bilge that LLMs can come out with I wouldn’t trust it to produce anything I wanted to put my name to.
Haringey Law Centre is not a normal firm of solicitors. It is a charity that provides legal advice. It proudly says on its website:
"As well as providing specialist legal advice, the Law Centre is active in our local community. We also work with sister Law Centres around the country to campaign on Access to Justice issues."
https://haringeylawcentre.org.uk/
As for Sarah Forey:
"Since being called to the Bar in 2021, she has effectively applied her legal expertise across various capacities that merge legislative compliance with active social intervention."
https://www.3boltcourt.com/team/sarah-forey-2021-second-six-pupil/
Make of all that what you will. I do think there is still substantial excellence in the legal profession, of which the judge's extremely witty incineration of Ms Forey is a great example!
No doubt there is - but it is diminishing, and tends to be found in older generations. A lot of the Tribunal decisions I read evidence a disturbing shoddiness across the piece.
Technology is not only an addiction, it is a way of life. The youngest generations feel they have to be on their phones to socialise and to keep up with the culture of their peers. I think a child who gets a phone later is one who will be confined to a much smaller group of friends.
It would also be a shame to miss out on content like this article, or some of the most educational YouTube videos through a dogmatic rejection of technology. But if I'm not going to reject technology completely, I have to contest with the diversity of content and its diversity of value. How do I know the content I am consuming was created by someone who has read first or second sources? Or someone who has condensed the information in a useful way? Or someone who has the professional integrity to write and research honestly? Articles like this one obviously rank higher than something on TikTok, but often the difference isn't so clear.
We categorise and teach forms of writing in school (expository, persuasive, narrative, and descriptive), I think it is also important to teach the differences in content quality and their value. This would be similar to the way some schools are teaching how to identify fake news. Once children are explicitly educated in the value of primary sources, educated writers and long form content, I suspect they may pay more attention to them.
Yes, I've noticed this. And I'm sure in the fulness of time we will figure out ways to better use/limit the tech. In the meantime, though, a whole generation has been very badly damaged
This is a problem in US legal practice as well. A lawyer was discovered to have used AI to write a court filing. The filing contained bogus citations. When confronted about the bogus citations, the law firm claimed to have cite checked the filing. Clearly the filing was not cite checked because cite checking is the process that was used to reveal the bogus citations.
https://www.businessinsider.com/claude-anthropic-legal-citation-lawyer-hallucination-copyright-case-lawsuit-2025-5
The problem is not limited to the law.
https://www.designrush.com/news/coca-cola-ai-backed-classic-ad-misquotes-ballard-misspells-shanghai
"Speaking paradoxically we may say that incompetence, having been standardized, has now become an essential part of professional excellence. We have no longer incompetent professionals, we have professionalized incompetence."
- Paul Feyerabend
Science in a Free Society, p 183
Very interesting read. However, I would hesitate to generalise too much from a case brought by Haringey Law Centre against Haringey Council. This is a poor, very diverse, London borough and I would be surprised if either party were able to attract any brilliant or ambitious lawyers of any stripe. It is, of course, shocking that this sort of farce is being conducted with tax-payer money.
Sure - and no doubt local government has always been characterised by a certain amount of laziness and sloppiness. But this is a whole other level, and is not what tends to go on in other countries (e.g. France, Japan).
Haringey is a very wealthy Borough. It includes what was said to be the most expensive street in Britain (Bishops Avenue) and upmarket neighbourhoods such as Crouch End. Even the areas with more social housing such as Tottenham are gentrified, being within a short Tube ride of the West End. Yet when I lived there, the two bottom placed high schools in the whole of the UK were both in Haringey. Its problems are entirely political, because it has been controlled by the hard left since the 1980s.
I share your pessimism of the future given the trajectory. It's a problem that isn't limited to law, unfortunately. Other authors, thinkers, professionals, and pedagogues have noted it as well. There's a very clear link between the death of the literary mind and the death of civilization. We saw something similar in the post-Roman era, when literacy and logic were at a low point. The digital world, by nature, precludes the human mind from long, complicated thought. The very same long and complicated thought that built the modern world. So I don't think it's a stretch to insinuate that the digital world will never produce another Einstein, Oppenheimer, Newton, Montesquieu, Lock, Nietzsche, or other great and world changing thinker (and no, I do not count the modern industrialists among those; I would trade 1000 Elon Musks or Steve Jobs for one more Einstein or one more Seneca).
Conversely, it's never been easier to be "above average" since the bar is now so dismally low.
Yes, the upside is there - it is not hard to stand out just by doing a little bit of hard work.