The Pope is concerned about AI impacting negatively on our humanity; the valuation of Anthropic rises to nearly one trillion dollars. It’s time to come out as an AI Luddite. I’ve worked in education my whole adult life, and it’s education that leads me to this position.
AI’s supporters say that it will make us all – including in education – more productive. A tiny minority of already extremely productive people conceivably could become even more prolific. Most of us will not. Our fate is that AI will make us increasingly lazy and intellectually deformed.
The trend is clear: soon in universities, faculty won’t write syllabi anymore. They won’t set essay questions and examinations. AI is doing much of this already. Soon it will do it all.
In their turn, students won’t be writing essays. At best, they will adapt the work that AI produces, with the sole aim of making it look less like AI has created it. But as AI will be grading the work, they won’t have to worry about even that.
Three years ago, as Vice-Chancellor at the University of Buckingham, I asked ChatGPT to write a graduation speech. It was nowhere near as good as the one I wrote unaided. To inject some humour into the proceedings, I included three paragraphs from AI in my speech, challenging parents and graduands to let me know in the drinks reception afterwards which paragraphs were not my own. Everyone got it right, that human writing being far superior then without a doubt.
Fast forward three years, in my final year as Vice-Chancellor, following the advice of several other university leaders who said it was what they were all doing now, I asked ChatGPT again. In a matter of seconds, it created a finely-honed speech that was much, much better than anything I could achieve through hours of laborious work. I gave it other tasks which were key to academic work – summarising important authors, highlighting relevant literature – and realised that it was leaving me far behind.
Then I knew we were in trouble. Much of what makes up the core of intellectual work – the ability to construct an argument, the ability to summarise, to précis, all abilities that I had thought were key to human engagement with the world – were now capable of being done better by AI.
So why not let it get on with this? Because that would mean our students losing these abilities forever, with faculty following not so far behind. Using AI is so intoxicating, students will no longer learn the basics of intellectual work and the underpinnings of our humanity.
For young people, becoming reliant on AI – as any user will become – is catastrophic. Letting ChatGPT do all the intellectual heavy lifting is like going to the gym and getting someone else to do, well, the heavy lifting (physical) for you instead. You could say, of course, the workout has been completed, the weights have been through all the motions, the necessary fitness steps have been achieved. Just not by you. There would be no benefit at all to your physical development. Likewise for intellectual development and AI.
What is required now is for a university to bite the bullet and ban AI altogether. Best of all, it should go back to physical books because in an AI-infected internet, these are now the only sources of knowledge that we can trust. Abandoning AI and going back to books is the only way now we can educate students in the best that has been thought and said, to better understand the meaning of our lives and our place in the universe, and to assist in the development of character.
Just as the University of Buckingham had to be created 50 years ago to address the problem emerging then, of overreach of government in higher education, there is now an urgent need for a new AI-liberated university to emerge.
It will need courage to bring this about. But parents and employers will soon see the difference between this new university’s graduates and those brought up on the false comfort of AI; it will be swamped with admissions.
The Luddites arose out of the first industrial revolution. Their inspiration was the – possibly apocryphal – Ned Ludd who, reportedly, smashed the textile machines in protest at how they were destroying the world his movement wished to conserve.
In education, we will smash the new ‘machines’, AI. We will create new universities, and parallel schools while we are at it, that will enhance human dignity in the time-honoured way, through individual intellectual and character development, not by getting an artificial intelligence to push us to one side. The neo-Luddites will not let our spirits be crushed or our intelligence made obsolete; maintaining our human dignity depends on it.
James Tooley was Vice-Chancellor of the University of Buckingham.


Discussion
Comments
This week across the site:
To join in with the discussion please make a donation to the Daily Sceptic.
Profanity and abuse will be removed and may lead to a permanent ban.
We let the rot set in decades ago by not banning computers, and Biros.
Reading list for the Pope: works by Adam Smith, F A Hayek, Joseph Schumpeter – the latter’s work on creative destruction being particularly informative.
Economic illiteracy is rife among the elites.
Great article, thanks. I like the example of going to the gym and getting someone else to do all the lifting.
All very concerning and will make most degrees worthless especially to an employer but also, ultimately, to the student.
The Luddites are already there in all university research departments. As an example, an AI assisted paper was produced which explained gravitation by a much simpler action than the conventional theoretical spin 2 graviton, which has been followed for the last 60 years. Various mathematical theories have been produced and Quantum Chromo Dynamics has been developed to a mind boggling extent where the vacuum of space is full of virtual particles, which are exitations of fields, flicker on and off and are never directly observable. This led to the biggest anomaly in physics where the total energy would be hundreds of times more than we observe through gravity. String theory produces 16 dimensions that we don’t see and even Sir Roger Penrose has admitted that he doubts it. Physicists have given up trying to explain how a graviton gets out of a black hole in order to have the greatest gravitational pull in the universe.
In 20 seconds, Claude adapted the latest research findings of a quark/antiquark virtual particle in the vacuum and the dual copy equations, which show almost identical forms for gluon scattering and GR, and suggested an action by which virtual photons, which mediate the EMF and… Read more »
So, AI is good at answering questions, but can it decide which new questions to ask?
It will soon.
No, it won’t. Not ever.
Excellent question, Jack the dog.
Was it the same AI which made this picture?
There won’t be any stopping or banning AI, any more than there was stopping and banning the introduction of any other technological innovation which made the skills of the working class obsolete in centuries and decades past.
I don’t doubt that AI will result in a de-skilled and dumbed-down population but they have already been de-skilled and dumbed-down by the “education” process in operation for the past few decades.
This sounds to me like a realisation and fear that AI will make our Universities and therefore their faculties obsolete.
Automate only those skills you’re happy to lose. I’m very reluctant to use it in any form. I like my mind and I want to keep it toned.
Same here.
Hear hear, me too.
I regularly use AI to generate BibTex citations. It’s much easier than surfing around in the hope of finding one that’s already been generated.
What is BibTex
I’m not entirely sure that citing the example of the Luddites helps his argument, with which I am broadly in agreement. Using AI is intoxicatingly easy. The output for a very minimal input gives one the very real impression of getting something for nothing, which is hard to resist, but not really sustainable.
AI is just super-fast computing. Calling it “intelligence” is both incorrect and misleading.
And – it is a tool. Tools need Human agency to be if any use. A hammer can’t knock in nails all by itself.
Oh yes, everybody KNOWS that you can stop bad things happening by banning them. Easy!
Murder was banned and it stopped happening immediat- ah, yes, I see.
It’s not necessary or good to ban it. Let them use it, let them cheat, then let them go into the world and see how they get on.
It is as you say.
The real test of knowledge isn’t an exam paper, it’s how or whether you can apply it.
In my early school days, during arithmetic the teacher would write some problems on the black board ( yes long ago ) but he would also write the answers.
His reason was he wanted to see how we got to the answer.
The suggestion of banning things gets my hackles up. I think we’ve had far too much of people wanting to ban things. How about persuading people that “AI” is just another tool, to be used judiciously, sensibly, with a desire to understand how it fits into the broader picture?
I used to think that about the internet; great servant that couldn’t become a master because all you had to do was unplug it or switch it off. Silly me.
You can switch some of it off, but other parts you need if you want to take part in the modern world e.g. online booking. Your point is a good one but banning things doesn’t work and I just don’t like the whole idea of banning things.
First, Woke needs exorcising from.academia.
Defund academia
Why do humans condemn themselves to obsolescence? Discuss.
Because we fool ourselves by calling it ‘progress’.
I suppose you would prefer to be in the fields scratching about growing parsnips and turnips for your supper?
We condemn ourselves to prosperity.
And you do not understand the concept of best allocation of labour resources. Labour replaced by a machine is free to be allocated to some other wealth creating endeavour.
For example: By making hand loom operators obsolete, many in the population could then afford things made of linen and cotton which previously were only for the rich, and thus created opportunities for new economic activity and jobs.
Thousands of farm workers were rendered obsolescent, replaced by machines. Food became more plentiful and cheaper – people were better nourished.. Labour moved from the land into industry to manufacture things people wanted and which made both the workers and consumers wealthier.
There is nothing here I disagree with except the ban because I have no idea how that could be achieved. I work in academic publishing s d we’re struggling to know how to detect the use of AI.
Oral exams maybe? You get first sight of the question on the day, you have thirty minutes to prepare notes and then you have to argue for your grade. You can’t fake what you never studied.