January 2, 2024 - 7:00am
On 1 January, the Catholic Church observes its “World Day of Peace” — a day which, since 1968, the Pope has used to address his followers on the prospects of peace in the year to come. The speeches tend to focus on a single issue, this year’s being of particular urgency both within and outside the church: the rise of artificial intelligence.
In the speech, the Pope spoke not only of the threats posed by AI in warfare and terrorism, but its potential to radically rewire society in a way that undermines human nature. Speaking in terms of regulation rather than complete condemnation, he asserted that AI “ought to serve our best human potential and our highest aspirations, not compete with them”. Central to his speech was the concept of limit: something, he remarked, that is “frequently overlooked in our current technocratic and efficiency-oriented mentality” and yet “decisive for personal and social development”.
He continued to elaborate on the dangers of this “technocratic” society, in particular the algorithmic manipulation of human behaviour — likely alluding to social media giants such as TikTok — and the threat of automation replacing jobs and diminishing the quality of labour. Both of these, he expressed, are dangerous because they pose a fundamental threat to human dignity — a dignity he was, of course, able to articulate in the theological language of our being made in the image of God.
For Catholics and non-Catholics alike, there is something to be said for the fact that the Pope is able to challenge AI with a coherent narrative about the world and our place within it. Arguably, one of the reasons why we are failing to regulate AI in the West is precisely because we are lacking such a narrative; beyond utilitarian explanations of material risk and financial gain or loss, few are able to explain precisely why human intelligence is qualitatively superior to artificial intelligence; why there is dignity in labour; and why technocracy imperils our quality of life. Without answers to these questions, we will surely be defeated by those who do have a vision of humanity — and those driving the AI revolution certainly do.
Since its inception, Silicon Valley has been motivated by an ideology based upon gnostic ideas, aspiring towards a transhumanist utopia that overcomes all limits. In reality, such a “utopia” — in making us reliant on technologies that only “the experts” can configure — will ultimately dehumanise us, creating a monopoly of resources and robbing ordinary people of the skills that make life truly dignified and meaningful. As C. S. Lewis prophetically remarked in the The Abolition of Man:
As such a scenario becomes ever more likely, the need to present a counter-narrative to the false utopias of the AI ideologues is vital — a counter-narrative that recognises the need to work with, rather than against, human nature. Such a vision is what Lewis called the “Tao”: the true and unchanging way for humans to live in the world, of which Christianity is an expression.
It is not enough to oppose AI out of a vague scepticism towards progress — those who do so are all too easily dismissed as “luddites”. Instead, those critical of AI need a vision that can properly account for human nature. Pope Francis’s speech was, perhaps, a testament to this. The sanctity of human intelligence and creativity is only tenable if we believe that we are made in the image of a creator; likewise, the language of “limit” is only meaningful in the context of a theology in which we have been given a designated place within the universe.
It is clear that the West will need such a vision if it wants to effectively mitigate the dangers of AI. Without it, the transhumanist paradigm will triumph, filling the void that secularism left behind.
Join the discussion
Join like minded readers that support our journalism by becoming a paid subscriber
To join the discussion in the comments, become a paid subscriber.
Join like minded readers that support our journalism, read unlimited articles and enjoy other subscriber-only benefits.
SubscribeThe religious perspective has to ring-fence humanity at all costs. Perhaps, though, we are no more than an expression of the complexity and richness of the structures within the fundamental particles and fields. Make the most of what time you have.
Assuming this fairly incoherent assumption is correct, then what *are* quantum fields? Are they temporal, and if so where did they come from? Or does time and space emerge from them, in which case what really can we conceive of this atemporal realm that follows consistent mathematical laws at any point in space, which can be a abstracted in human terms that ate not too simple or too complex for us to understand something of them such that we can make predictions? What is maths itself and where does that come from? How can we explore this realm and find objects like imaginary numbers that then turn up in nature?
I assume you don’t act as if nothing has meaning, that no philosophy or ideology is really true, just a useful construction, that murder is only wrong because evolutionary psychology has shown that to be a useful assumption. If you can get away with it, there is nothing objectively wrong with murder. Do you assume that the ontological horizontal is causally closed, that you have no free will and that the plays of Shakespeare were effectively written at the point of the big bang?
If your answer to any of these is that you don’t go around like a hopeless zombie unable to have any real agency, values or meaning in the world, then maybe you can appeal to contrived meta morals or magic compatibilism etc, but you are appealing to some very strange type of mindless magic.
I think you missed the point of Kathleen’s “perhaps”.
Couldn’t you just have said that we are more than just ‘stuff’, and saved us from that torrent of bollocks? The ‘ontological horizontal is causally closed’ is surely one of those ‘reach for your pistol’ moments. Bravo!
The Pope’s recent statements seem to be getting more and more panicky – as if the Catholic Church just can’t compete with life.
As an old person I can see how the Pope (even older) wants to guide young people away from his view of a disastrous future. I had many lectures from my grandfather along the same lines – if it’s new treat it with suspicion!!!
And the effect it had on me? It sounded like something to try, quickly. Which children (with any spirit or drive) listen to their grandfathers?
Many children listen to their grandparents. It’s their parents that they don’t listen to.
Depends on the parents, depends on the grandparents – depends on you!
Indeed. I was generalising from experience.
Sorry, I’m not taking this narrative about limits (“stay in your lane !”) from a Catholic church which claims that its “truth” has no limits and that its leader is himself infallible. One rule for the rulers, another one for the plebs.
Meanwhile, out in the real world, science and technology continue to advance unhindered by such ideologies, utopias and fixed beliefs. Whatever is actually is, regardless of whether it fits our current belief system. The belief system adapts pragmatically to absord our better understanding of reality (which has changed enormously over the past 2000 years as Pope Franics might care to stop and reflect on).
“Since its inception, Silicon Valley has been motivated by an ideology based upon gnostic ideas, aspiring towards a transhumanist utopia that overcomes all limits.”
That’s completely untrue. Esme Partridge is too young to know anything of the history of Silicon Valley and has almost certainly spent little time (if any at all) there. As with any society, there is a spread of views and ideas. The concept that everyone in Silicon Valley believes in the same utopian ideals is nonsense. Indeed, engineers are by nature pragmatic people focused on problem solving. They have to be. They get paid by results and if stuff doesn’t work, they don’t get paid (which isn’t always the case for commentators).
Again, this feels like projection: a committed ideologist claiming that her opponents are the ideologists when they are in fact less idealogical.
I note also from the tone of the article that the author has an in-built prejudice against AI (“It is not enough to oppose AI out of a vague scepticism towards progress”). Why exactly ? Any technological change brings both benefits and costs, winners and losers. That’s always been the case. The challenge for us is to maximise the gains and minimise the losses. Not to gnash and wail.
This won’t be the worst UnHerd article of 2024. But it’s not a promising start.
I don’t think you understood it
I don’t think so.
AI is being opposed here not because it is “dangerous”, but because it challenges embedded beliefs like “dignity in labour”. The hyped-up “danger” is merely a cover for the ideological struggle (which is, as I stated, against the pragmatists).
What exactly is “dignity in labour” anyway ? It’s an entirely human invention which has no real basis – pure ideology then.
Likewise this: “The sanctity of human intelligence and creativity is only tenable if we believe that we are made in the image of a creator”.
It is *assumed* in the article that there is something special and unique about human intelligence. That’s a belief, not a fact. She states that human intelligence is in some way that’s hard to explain “qualitatively superior” to AI (but doesn’t explain how this is). If so, does the same hold for human intelligence vs animal intelligence ?
I suggest that over the next century we shall discover yet more things about the actual nature of scientific reality that contradict many firmly held religious and political beliefs. Things like the extent to which intellgence is heritable (which is probably above 0%). If that’s the case, perhaps we can stop wasting time on things like “equity”and get on with solving real problems.
Learning a trade or a craft and participating productively within a connected community can be a source of great meaning. ‘Dignity in labour.’ It’s pretty basic common sense stuff.
But you don’t seem to understand. It’s what many people are lacking in a current virtualized, anonymised, standardized, bureaucratised, de-skilled and abstracted services-leaning economy. Trends which AI will perhaps intensify.
Traditional worldviews might provide a helpful recalibration towards human nature, in this context – and away from the blank slatism of Silicon Valley, or the radical change of constantly mutating technological developments, and tech centred occupations, or the idea that human beings are just fungible components within a system, etc.
But I do understand that. I rebuild old bicycles as a hobby and there’s a fantastic community of like-minded people involved in this. It’s very rewarding and perhaps more so than anything I’ve done as a job.
But there is no requirement that a job must offer me that level of satisfaction. Nor any way that AI is going to prevent me tinkering with dirty, broken old bikes as a continuing leisure activity. If AI gives me more leisure time to engage in hobbies, that’s a result for me.
Human work started out as hunter-gatherer survival activity. If there was any job satisfaction or dignity in that it was by accident rather than design.
In fact, put properly to use, AI can eliminate much of the boring, repetitive work people do not find rewarding. Are we in danger of fighting to preserve current work patterns purely because we are by nature frightened of the future (which we all are to some extent) and lack the imagination to see how things could be better ? Many people wanted to preserve coal mining jobs in 1984. Most of those people now actively oppose new coal mines – even if the jobs now would be far safer and better than they would have been in the 1980s or earlier.
You’ve digressed. I was explaining what ‘dignity in labour means’ and was thinking more along the lines of professional workers embedded in a meaningful matrix of economic production – eg. industrial workers in industrial towns/ craftsmen/ tradesmen, people professionally serving their communities, rather than just private hobbies and pursuits.
Meaningless work is a problem, which will probably get worse as AI transforms the economy.
Correctly applied, AI will eliminate most of what you call “meaningless work”. It is the lack of “meaningless work” which in many people’s view will become “the problem”. It’s not a problem. It’s an opportunity.
I think you are ignoring the real opportunities that AI- like any new technology – offers.
Meanwhile, actual hands-on jobs like electricians and plumbers and people providing personalised local services will continue – and actually thrive. And this is – in my view – a good thing.
I suspect we are actually after the same ends and just disagree about whether AI will help or hinder those goals.
And some people will turn their private hobbies and pursuits into income-generating activities. We’re just going to have to get beyond the assumption that we all have lifetime secure full time jobs (which many people do not enjoy anyway).
All slightly naive faith-in-the-future stuff, which ignores the extent that current tech has already atomized & hollowed out current society.
Yes, a limited use of AI might be helpful and provide some benefits. But general knock on effects, social disruption and negative externalities from widespread use are difficult to predict and might be wide-ranging. And if AGI is possible, embodying a runaway super-intelligence, then who knows what that entails.
I attended a conference in California and spent lunch with a medical doctor who advised the Obama administration on health-and-social policy (c2017).
They estimated over 20% of all adult males in employment worked in some form or other in ‘transport ‘. They reconned making transport functions entirely automated was likely in the next 10-15 years.
The social consequences of such massive male unemployment would, they thought, turn the US in a post apocalyptic nightmare.
It’s not only transport.
My daughter’s best friend is a super-talented linguist who is now part-time editing- in the full knowledge she’s really training an AI bot how to replace her.
Machine learned bots CAN and DO read X rays and scans better than humans.
I’m not saying these things aren’t better (I’m sure a bot could drive more safely than me), but the societal consequences for everyone( because the technocrats of today are simply developing THEIR own replacements) are world-changing.
And not necessarily to mankind’s advantage.
… indeed, probably the direct opposite.
Blithe optimism is naive.
I’m not much of a papal fan, but he’s nailed this one.
All he is really talking about is setting limits. After global pollution across air, sea and land, ongoing increases in depression and suicides, habitat and diversity destruction to the point it may be unsustainable, and we have already lost vast quantities of biodiversity we will never get back, are you really saying that discussing sensible limits is somehow anti progress?
No, I am not.
However, he should recognise that technological advance has also reduced pollution. For example, car engines today are twice as economical as they were 40 years ago and far less polluting. The River Thames was far more polluted in London 50 years ago. My parents and grandparents grew up with pollution and smog far worse than today in West London suburbs.
He has chose only to emphasise the negative possibilities of technology.
Take his mention of “and the threat of automation replacing jobs and diminishing the quality of labour”.
Where is the consideration that automation not only destroys older jobs, but also creates new jobs ? That has always been the case in the past. They are just different jobs. And often better ones. Where does he consider that the “quality of labour” (a term I doubt he’s ever defined) may be improved by automation and technology ?
To which i’d add her assertion that:
This is simply nonsense. The concept of “a creator” arose just a few thousand years ago, and coalesced into the major religions that we’re now left with as a legacy. So how did human intelligence and creativity not only survive but thrive in the many millennia prior to these recent developments? It didn’t require “sanctity” but the development of social engagement once the agrarian revolution got under way. The whole concept of “sanctity” is a form of deceit designed to induce compliance, but it just won’t wash any more.
Humans are spiritual creatures, which arises from our consciousness of ourselves. I can appreciate that my neighbour has that same consciousness, and would therefore seek to defend my neighbour as a means of helping secure my own defence against external threats. Mutuality is all that’s required, however complex that becomes.
Edit: just read your later reply on the same quote.
If an Ai is better at finding a tumor in my medical scan than a bored, overstressed medtech, then to hell with “dignity in labor.” Bring on any AIs that can be trained to do given tasks better than humans can. I’m hoping that within my lifetime this applies to driving, too.
I’m impressed that the Catholic Church is, in my observation, the first to react to the rise of AI with a sound understanding of potential dangers and a cogent framework to deal with it. I expect, what we will instead have in the (English speaking) West is AI will break the societal order (for example by making the vast number of jobs redundant such as many drivers, cooks, baristas, translators, lawyers, sex workers, even tech workers themselves as well as manipulating human behaviour according to the wishes of powers that be) and cause large scale destruction before the idea of limiting it will go mainstream. Since watching it as a child I was intrigued by why Blade Runner (the original good one) was such a nihilistic and negative depiction of what appeared to be a wonderfully technologically advanced society which seemed to have everything. I’m coming to the realisation how prescient they were at the time perhaps also adding the likes of Robocop and the Terminator.
The Pope has many good points, as he generally does. He reminds us that just because we can do a thing, it does not follow that we should. The concepts of caution, moderation, limitation, and setting appropriate boundaries are important in many phases of life, not only technology. He encourages introspection, self-criticism, and humility, traits that are much needed but in short supply these days.
That said, the apocalyptic and/or messianic possibilities of AI are being built up to such a degree that I wonder if it can possibly fulfill such lofty expectations. Supporters are projecting their hopes onto AI and detractors are projecting their fears onto it. Both images are irrational and most probably wrong. I don’t think AI will destroy us, save us, or ever entirely define us. I think it’s far more likely that people trying to use AI to usher in a utopian future, or using it as a bogeyman for imagined threats will cause most of the problems. I think the truth is probably a great deal more mundane, that AI, like other technologies, is a tool like others that can be misused, abused, and overused to the detriment of all, but like other things, we’ll have to discover the pitfalls as we go.
Pope Francis is plainly not a coherent thinker. This is the guy who signed the Document on Human Fraternity in Feb 2019. Among much other drivel, it proclaimed that all religions are willed by God. Amazingly, his cosignatory Grand Imam al-Tayyeb, as a good Muslim, had earlier proclaimed that unrepentant apostates should be executed. So I see little reason to take any notice of his warnings about subjects outside his tiny area of expertise.
The current limits of AI have already been horribly evident in the deaths caused by self-driving vehicles. Other limits will doubtless become evident as overambitious experimenters try automating other areas of human endeavour. But I do not doubt that there will be huge benefits if more routine jobs are automated or receive automated assistance.
I am being dense here, but can someone explain to me why AI is such a danger to society? Because it can write (some kind of) sermons to priests?
I read everywhere how weary we should be of AI, but the message is not getting across, as far as I am concerned, as to why.
One obvious reason would be the elimination of jobs – in their millions. AI doesn’t need to be as good as a human; just cheaper. “Not quite as good as a human” is good enough for business. And wages will in turn be driven further down for those who can hang on to their jobs. Right across the workforce, jobs have been dumbed/down and de-skilled already, so that human workers can be trained fast, paid poorly and swapped out quickly. It makes sense and it’s inevitable from a business point of view, I suppose.
It won’t remain good enough forever. AI will take over humans for most tasks including driving safely, writing/reading, translating etc etc. There’s no limitation other than computational complexity and power that limits the analytical ability of AI. Today we have 70B parameter LLMs, in a year or two it’ll be double that. This is leaving aside more powerful reinforcement learning models which will probably blow away LLMs in their ability when they arrive.
But what kind of jobs?
No, I cannot.
It is, however, a threat to the entrenched power and privilege of large groups of people who will therefore do everything they can to oppose it.
It is basically just another new technology that, if correctly harnessed, offers huge benefits to humanity as a whole.
In some ways, it is finally bringing automation and greater division of labour to activities which have up to now been immune to this (things like journalism, writing, music, law) – as well as many technical fields. In these areas, there has traditionally been a large baseload of more routine work which has been hard to automate and a tier of highly skilled, difficult work at the top which will remain a human field. Anyone in these fields needs to make sure they’re in the top 15% and figure out how to harness AI to become even more productive.
This is no different in nature from the automation that’s worked through manufacturing over the last 150 years. It’s just a different – and far more vocal – group of people complaining about it.
Sceptics might pause to reflect that we’re not too far away from being able to real time check the work of the so-called “fact checkers” (think the BBC’s Marianna Spring) and expose their distortions … and even automate away her “job”. That will be thanks to AI.
Imagine a real-time BS/lie detector running in the House of Commons which produces a loud beep every time some lie/misinformation/distortion was said. It would hardly stop beeping. Perhaps the politicians might then have to up their game.
Why is everyone looking/hoping for the worst ? What a way to start the New Year. (And I’m not by nature one of life’s optimists).
All fine as long as you’ve no ideological objection to UBI. I can understand why you might revel in the prospect of those in the groups you identify being kicked out on their *rses, but better to have plain folk (like me) spending their days doing things, in my view. You know, “idle hands” and all that. AI will spawn very few new jobs. Jobs are a very unwelcome by product of the pursuit of wealth.
Every time automation has been introduced before, many people believed that it would create permanent mass unemployment. But so far that has never been the case. Why is it any different this time ?
I detest UBI. It’s a dreadful idea which divorces income and reward from effort and achievement. It’s also unfunded. Besisdes which, the rich will simply emigrate as they did in the UK in the 1960s and 1970s.
It’s funny the resistance of oldies to UBI. Do they really believe effort and income are fairly linked? Are poorly paid nurses not putting in huge effort? Are [removed ‘below’ in edit 11:15 3/1/24] minimum wage factory workers at Amazon warehouses (pissing in bottles to keep the conveyor belts running) not putting in huge effort? Arguably the insane asset bubbles, especially in property, has been a form of UBI for Boomers. It’s allowed a whole generation to do what they want without much effort.
I never said they were perfectly or fairly linked. But there certainly is some linkage – which is far, far better than none at all.
UBI removes a core incentive for people to learn skills and improve their competence and productivity. Why bother training to be a doctor, lawyer or engineer ? But society still needs these people. The whole idea is beyond stupid.
Are nurses really “poorly paid” ? Getting rid of national pay scales and providing market weightings outside Greater London (e.g. in the South East) might be a start in sorting out some of the problems here. Nor are they forced to be nurses or stay in the NHS or the UK if they feel undervalued. Indeed, I believe there’s a thriving markey in agency nursing for some of these reasons. The fact that the NHS is poorly structured and managed is however nothing to do with AI … though AI might in due course improve medical efficiency and productivity which has hardly moved despite the massive increases in computing power, productivity and automation in most other activities over the past few decades.
The Amazon warehouse jobs will be automated. Amazon has always been clear that this is the plan.
If you have evidence that Amazon workers are paid below minimum wage, perhaps you would a) show it here and b) report it to the authorities so it can (as it should be) be prosecuted. If you don’t, you should withdraw this.
I agree with you on the insane asset bubbles and the social damage these have created. I’m not sure that’s a result of technological change. I’d lay this at the door of governments (allowing execessive immigration whilst restricting house building and desperately underwriting the inflated “housing market”) and central banks (creating too much cheap money).
And yet you learned to repair old bicycles and take great joy and satisfaction from it. You didn’t do this to earn a living, you just did it. This old mantra that no one will be motivated to do anything worthwhile, unless they feel the terror of an empty belly, is clearly false.
UBI will allow people to do things that are worthwhile instead of the mind-numbing nonsense that make up 90% of jobs. Will some people sit around endlessly drinking and w@£king? Yes. But far less than you think.
Great comments Peter B- as you allude to- this is a technological forest fire, that will burn down the swathes of skeletal dead trees, allowed to stand leafless too long, their calcified roots gripping exhausted soil at the expense of new growth.
I should point out here that I – as an aging engineer – am not immune to the techno-anxiety that seems to underly so many of the comments about AI.
One year ago, I was rather daunted to discover just how far technology in things like code scripting had moved on over 20 years in which I’d quite successfully continued polishing my legacy softweare scripting skills, oblivious to the new and more productive ways of doing things. How could I ever catch up ?
But after a little time learning about more recent languages like Python, it quickly became obvious that things that has taken me far longer to create in the past were now easily achievable in a fraction of the time and with a fraction of the code. Who wouldn’t want to have such fantastic tools at their disposal ?
And was I really at a disadvantage to the 20 year olds who already know this stuff ? Not at all. In fact, I think the reverse. Knowing how to work with older tools and techniques and limited resources and the discipline this requires is actually a highly valuable skill. All these AI systems need people to build them and check they are doing what they should. The demand for highly skilled engineers isn’t going away – it’s only increasing.
We can talk on endlessly about “setting limits” and “regulating algorithms”. That’s not going to happen. This stuff will happen now. We just need to make the best of it. If you’re that worried about it, be one of the good guys and get hands-on building and checking these systems. There won’t be any shortage of work here – and it will be interesting and well paid.
Another mystery downvote…
Can you give me an example of a job in the writing, music or law where this could be handy?
I suppose things like conveyancing which is boring in the extreme and repetitive could easily be automated.
Slightly tangential, but I was talking to a friend yesterday about AI and he noted that he used AI to create bid proposals (this involves some legal stuff).
His point – which I agree with – is that AI can very quickly get you to a “90% solution” – the basic framework and most of the text. He/his team then clean this up to the finished proposal. This saves a lot of time. It also frees up time to work on morecreative aspects of the proposal rather than the mechanical aspects of putting the proposal together.
None of that helps if you’re clueless about bid proposals and don’t understand your customer requirements. So I don’t see how that’s “deskilling” the job. But it helps a lot if you do.
In visual design you can already create AI artworks. Again, I can see ways this would provide a useful starting point for further customisation. For people like me who lack artistic drawing and painting skills, it also means we can now more easily create artworks provided we have the some conceptual goal. We are less limited by technical limits. Why should artists need to be both masters of painting technique and imagination ? (Of course, there are some relationships between these). But I don’t see how this progress is a “bad thing”.
This is why abstraction in the creation of art has become necessary, because those with the most “talent” at reproducing the external world weren’t necessarily the most talented in terms of artistic imagination.
It’s my opinion that AI might be able to produce very pleasant artworks, but without a heart, viscera and neural contact with the outside world, it’s products will remain of secondary importance to those produced by human beings.
I agree.
It is the “finishing layer” on top of the “90% good enough” base that will be the difference between a mass-produced image and one which sells for much more.
If this is true, it means that yet another human activity is moving towards a “winner takes all” scenario – much as has happened with international pop starts (Taylor Swift is a billionaire), top footballers, giant tech companies, …
I do not however see this as being inherently “wrong”. At the end of the day, this small group of people benefit because they are able to provide what people want better and/or cheaper (usually better, often more expensive) than their predecessors.
The challenge is in sharing the wealth created in a more sensible way – and not in trying to stop the inevitable. Taxing the giant tech companies properly (and closing their huge tax loopholes) might be a start. People need to get off their backsides and start clling companies like Apple tax cheats rather than naively believing they are somehow the good guys.
Why the downvotes???