ChatGPT: What No A-I Hath Seen

By: Stephen McAlpine

The UK TV series HUMANS (the A” is upside down) from a few years ago was a cracking insight into the world of AI, in a near future UK setting.

Humanoids are synthetics or “synths” as they are known, and they have been created that are remarkably real, and they do the bulk of the manual work for the planet. But not just the manual work. They are programmed to grow relationally and emotionally they more the interact with humans. Or they seem to do so at least. They’re baby-sitters, colleagues, and, as we shall see, allied health clinicians.

The synths are hard to pick from real humans – hence the title of the series – and are designed to smooth out life for the average harassed and harried middle income family just trying to get by. As in all great Sci-Fi, the future is easy to imagine – or fear -, but it lies just beyond our current reach.

Some things don’t change though. Relationship struggles between real humans can’t be “botted” away. Having your own synth comes with the promise that you can get on with more of your increasingly self-actualising life, and leave the drudgery behind. Yet none of that seems to account for the frailty of the human condition, a frailty not seemingly exhibited by the synths. One of the couples in the series have been having ongoing issues, exacerbated by the husband’s decision to have sex with one of the AI humans, all in the selfish and naive notion that it wouldn’t be real sex or real adultery.

But as these things go, seeing an actual counsellor was proving hard, with a long waiting list. Instead they commit to a series of counselling sessions with a AI Human. And I mean, it’s hard to tell. The answers seem the same, the philosophy of life seems to be taken straight from the psychology Bible, the DSMV. All of the wisdom emanating from the “synth” – styled in both dress sense and office decor as any self-respecting clinical psychologist would – is the collective wisdom of thousands of clinical hours. Yet something is missing. The couple in question go for one session, then never go back, intent as they are in sorting out their flesh and blood troubles with flesh and blood.

Computer Generated Blogging

Now I’m a late adopter to technology, so when I first starting seeing computer generated blogging advertised on my Facebook feed, I was “Seriously? For a start, the whole point of writing, and my blogs in particular, is that predictability means death. It’s the road bumps on the road in good writing that makes it work. It’s putting together ideas or concepts that are seemingly paradoxical, or clunky, and making them work.

Yet here we have elite private schools in the UK stating that the era of essays and exams doing on computers is over, as ChatGPT has been able to write essays that get an A+. The reason of course is that our education system is not looking for creativity, but uniformity.

In a provocatively titled article in The Atlantic, “The College Essay is Dead“, Stephen Marche pitches the humanists and the scientists against each other, observing that:

“The essay, in particular the undergraduate essay, has been the center of humanistic pedagogy for generations. It is the way we teach children how to research, think, and write. That entire tradition is about to be disrupted from the ground up…The extraordinary ignorance on questions of society and history displayed by the men and women reshaping society and history has been the defining feature of the social-media era.”

Humanities teachers and posh schools can bang on all they like about how their role is to teach students how to think, but in the end the system is interested in results. Besides, life does not generally reward you – at least not in the David Foster Wallace “trenches of adult life” – for thinking too much about anything. And certainly not for truly critical thinking that disturbs the status quo. Humanities departments of universities are not populated by free thinkers, but by group thinkers.

ChatGPT Maintains the Status Quo

The great irony being that they are generally oblivious to this. Yet post an essay to your critical theory tutor at university that’s actually critical of critical theory, and wait to see the groupthink kick in. Perhaps the draw of ChatGPT and the like by humanities students is that they know the system is gamed. Original thought is out the door. Peer review has seen to that. The idea of peer review was supposed to be that quality would be maintained. The reality is that the status quo is maintained. ChatGPT will ensure that reality.

What’s more, life is so busy, jobs are so tenuous and relationships so fraught, that true freedom of thought will paint you into a very lonely corner. Truly free thinking comes at a cost.

And what about theology? Beyond a theological essay that’s required at college? Nathan Campbell’s attempt at creating a “church plant” with ChatGPT was enlightening. And chilling. Nathan came up with a myriad number of church plants, with any number of funky names, vision statements, music lists, website ideas, design layouts for buildings, etc. Truly amazing. I recommend reading Nathan’s whole blog post. In true Nathan style it’s long, but it’s worth it.

But as I said, it’s truly chilling. The churches that ChatGPT would come up with would slip down like smooth peanut butter, there’s nothing crunchy about them. In a culture absolutely committed to safety, these churches would pass the “Safe Churches” test with flying colours. At a church planting conference there’s the chance such a church would be given an A+, and used as an example of what to do. It seems such a world away from my late teen years in the Assemblies of God churches in working class suburbs of Perth, in which God seemed to just turn up at unexpected times and “beat up” on us, or

In fact they just might be already. Nothing in Nathan’s blog post examples seemed all that different to many of the churches that dot the landscape of any modern Western city. All we need now are preachers who are “synths” and the job will be complete. The only burnout a pastor will experience in the future will be when something electrical short-circuits.

Church in a Box Comes With Risks

Nathan does illustrate however, that such a technologically fuelled church in a box comes with some risks – or at least ChatGPT thinks so. Risks include a certain alienation that something too smooth and seamless would engender in the messiness of real life. Something faux in the manner of the synth marriage counsellor in HUMANS.

Yet that’s not going to be ChatGPT’s conclusion without it being asked a leading question. You’d have to ask “What are the risks?” before it would conclude that you want it to list the risks. You’d have to aware that there are indeed risks to a safe, smooth, vanilla expression of church.

You can imagine the tribes of Northern Israel in the times of Elijah asking ChatGPT “What are the benefits of Baal worship?” and being given a wonderful list that includes flourishing crops and bouncing baby boys aplenty. ChatGPT is not about disrupting the flow – it’s about removing the barriers to where you wish to go. Assuming of course, where you wish to go is the right direction.

Now it’s not all naysayers. As usual there are those who see a ChatGPT future so bright that we’ll need to wear shades. As reported in The Times recently:

Richard Susskind, co-author of the book The Future of the Professions, said:

“Almost everything that we do today online will be more convenient because of this, booking the cinema, buying clothes and so on. But my excitement, in many ways, is not about enhancing our existing daily uses of technology. It’s about how we have, in our guts, the sense that we can solve the global health problem, we can solve the global education problem, we can solve the global justice problem by somehow having different ways of spreading the expertise.”

I get the convenience thing, even if I constantly refuse the predictive texting in my emails, just to ensure I’ve got my own brain switched on and going down the neural pathways I wish it go down. Yet is it true that we’ll solve the global health problem, the global education problem or the global justice problem? Are those problems about lack of knowledge per se, or are they just as readily about lack of will, and lack of philosophical alliance?

Do you believe, for instance, that a major global health problem, as you see it, is not enough access to abortion services for women in third world countries? Or do you believe that a major global health problem is the termination of so many viable foetuses? How many tweaks will it take in our AI development to launder out the responses, and to craft a problem that was once seen as a solution?

One person’s terrorist will, I presume, remain another person’s freedom fighter, despite the advances of AI. The likes of Susskind seem to have this breezy optimism that we’re all on the same page, and when eventually all of the information is literally on the same page, we’ll somehow all agree with it. Fine if we want a smoothed out life, which on the surface of it, we probably do.

Stephen Marche in The Atlantic observes

“Contemporary academia engages, more or less permanently, in self-critique on any and every front it can imagine. In a tech-centred world, language matters, voice and style matter, the study of eloquence matters, history matters, ethical systems matter. But the situation requires humanists to explain why they matter, not constantly undermine their own intellectual foundations. The humanities promise students a journey to an irrelevant, self-consuming future; then they wonder why their enrolments are collapsing.”

In other words, we’re combining peak nihilism with peak efficiency. Sounds like a recipe for disaster. Which it is proving to be.

Of course the one human that AI or ChatGPT could not replicate or predict would be Jesus, right? Because we just couldn’t predict the mind of God who would come up with the gospel message. Sure, if you were to ask ChatGPT what the gospel is, you’re probably going to get the right answer in there (hence the chances of it being used to write theology essays!). But to ask ChatGPT to come up with the gospel? Got as much chance as a real human doing that.

How did Spurgeon put it?:

“I am sure it is God’s Gospel; for nobody could have invented it. A plan so just to God, so safe to man; and I am all the more sure it is God’s Gospel because there are many that hate it.”

Which is true of the church too. We just couldn’t come up with it – in the sense create it out of nothing through the resurrection power of the Holy Spirit for the purposes for which God intended it. No AI could come up with that bunch of people in that place doing that thing. The late, great Marva Dawn wrote a great book about public worship called A Royal Waste of Time: The Splendor of Worshiping God and Being Church For the World. She wrote it back in 1999, on the cusp of the new millennium, one that has begun apace with technological advances even she didn’t see. She didn’t see the full flowering efficiency that the church is becoming, at a time when inefficiency is frowned up – or photoshopped.

Yet what do the Scriptures say?:

Now I rejoice in what I am suffering for you, and I fill up in my flesh what is still lacking in regard to Christ’s afflictions, for the sake of his body, which is the church.  I have become its servant by the commission God gave me to present to you the word of God in its fullness – the mystery that has been kept hidden for ages and generations, but is now disclosed to the Lord’s people. To them God has chosen to make known among the Gentiles the glorious riches of this mystery, which is Christ in you, the hope of glory.” (Colossians 1:24 )

There’s just something wonderfully rambunctious about that. And I like the fact that AI would never predict a blog post putting the world “rambunctious” after the word “wonderfully”. Imagine how “cray cray” Gentile life once was, so riven by suspicion and division. Imagine how the Gentiles, having exchanged the glory of the Immortal for the glory of created things. And then imagine the gospel coming into that situation, and upsetting it all, sweeping away the smooth quid pro quo of pagan worship life, and injecting gospel worship into the mix.

Or to put it another paraphrased way with thanks to Isaiah 64 and 1Corinthians 2: “What no A-I hath seen, what no ear has heard…”

Don’t take me for a Luddite. I said I’m a late adopter of technology, not a rejector of it. So there’ll probably be plenty in this blog that I’ve missed out on, or misunderstood. As blog posts go, it may only be a B-. But I like my church a little rough around the edges. I don’t want it to be A+ every time. if it were I might be content to settle for the church as it is this age and not as it will be in the age to come. An A+ church will require A+ people, a synthetic church comprising the collated pro-forma of what every pastor is looking for in a congregant. After all, who would want a church full of synths when you can have one full of humans?


Article supplied with thanks to Stephen McAlpine

About the Author: Stephen has been reading, writing and reflecting ever since he can remember. He is the lead pastor of Providence Church Midland, and in his writing dabbles in a number of fields, notably theology and culture. Stephen and his family live in Perth’s eastern suburbs, where his wife Jill runs a clinical psychology practice.

Feature image: Photo by Andrea De Santis on Unsplash