Through a consideration of examples of the AI Creation Meme, a remix of Michelangelo’s Creazione di Adamo featuring a human hand and a machine hand nearly touching, fingertip to fingertip, this article will tackle the religious continuities and resonances that still emerge in AI discourse in an allegedly ‘secular age’. The AI Creation Meme, as a highly visible cultural artefact appearing in a variety of forms and locations, will be analyzed and discussed for its religious, apocalyptic, and post-humanist narratives, along with reference to earlier work on the New Visibility of Religion—specifically, Alexander Darius Ornella’s consideration of the New Visibility of Religion and religious imagery of the 2006 film, Children of Men. Work that outlines the aspects of critical post-humanism, speculative post-humanism, and transhumanism in relation to the contemporary post-secular age will also be addressed to expand on the implicit apocalyptic messages of the AI Creation Meme. Such a consideration of repeating and remixed imagery will add to the scholarly conversation around AI narratives and the entanglements of religion and technology in our imaginaries of the future.
This post will include spoilers for Devs, a new television series by Alex Garland (2020) and for Tom Stoppard’s play Arcadia (1993). Although, really, you should have had plenty of time to catch the latter by now… Anyway, spoilers ahead.
Seriously, spoilers. You have been warned.
You should turn back now if you haven’t seen Devs and Arcadia.
Seriously, you can turn back now…
And the fact that you can turn back is actually significant [spoilers start from here].
Because you can, if you choose to, turn back here. You are able to. You can click back on your browser, and reverse the flow of time in a metaphorical, if not a literal, way. Here, at this moment, you can go backwards. At this moment, you CAN stir the jam back out of the rice pudding…
“When you stir your rice pudding, Septimus, the spoonful of jam spreads itself round making red trails like the picture of a meteor in my astronomical atlas. But if you stir backwards, the jam will not come together again. Indeed, the pudding does not notice and continues to turn pink just as before. Do you think this is odd?” – Thomasina, Arcadia
I’ll get back to Thomasina’s observation in a moment.
I recently finished binge-watching Devs by Alex Garland. As television series go, I’m not sure that Mr Garland could have created a story that was more up my street than this one. Yes, like almost all the other AI thinkers, researchers and technologists I’ve ever met, I’ve seen Ex Machina. But to be honest, I probably liked it more for its take on Silicon Valley bro-preneurs and bro-tagonism than for what it said about the probability and nature of AI consciousness. And some of the male-gaze moments felt more like having your cake and eating it than critique to me. Although, the alternative ending where we got Ava’s non-human POV always appealed. Annihilation scratched a few bio-horror and posthumanism itches and gave me strong female scientists, and gets a thumbs up from me. So, I was interested to see what was coming next from this writer/director…
Not only Silicon Valley culture but also Simulation Theory, computational universes, determinism, multiverse theory, and all replete with religious imagery and narratives. Because of this near-perfect mix of favourite Beth-things, I’m going to forgive rather the on-the-nose last episode twist of the V in Devs being a roman numeral making it actually the U in Deus. Computers as god(s) isn’t a new story, and I’ve spent a lot of time writing about theistic conceptions of AI. But AI wasn’t really the focus here. It was more about the human characters than the tech in the end. In fact, the first episode took us quickly from AI to Quantum Computing as one of the key characters was promoted from the former to the ultra-secretive DEVS, indicating that this wasn’t Ex Machina 2.0, even if we finally got the ‘Deus’ part of that old expression.
Even the lovely shiny golden Quantum Computer in the floating box under the ground surrounded by a vacuum (how DID the toilets work???) wasn’t much more than a McGuffin for a very human story about fate, free will, and sadness. Yes, the story wouldn’t have happened without the computational power of this near-mythical golden device; resplendent in shining doodahs, tubes, and wires, and with a pulsing heartbeat that shone and shimmered around the underground cube. Deus was able to crunch such impossibly large amounts of data – all the data in our universe in fact – that both the past and the future could be predicted and projected visually on a screen. Seeing a static blurred Jesus on the screen in the Devs team’s red screening room in the floating box was a powerful moment, even for a non-believer. Hearing him speaking in Aramaic was genuinely spooky. But Devs had bigger aims than dealing with the implications of finally proving the historical existence of Jesus.
Because, seconds later, in a heart-rending moment of crushed dreams, the developer, Lyndon, who had finally managed to fix the white static fuzziness and produce crystal clear audio of those last words, was fired. He had introduced Hugh Everett’s Many Worlds Interpretation into the Quantum Computer’s prediction algorithm. Forest, the head of Dev’s parent company Amaya, had banned Many Worlds theory because he was dedicated to a deterministic view of a single universe rather than the branching possibilities of many universes as described by Everett. Introducing multiple worlds meant that they might not be predicting their Jesus, for example, but one who might be from another universe. Different perhaps by only by a single hair on his head, but just not the same one.
But it wasn’t Jesus that Forest was interested in. And given his later comments about the Messiah complexes of Silicon Valley genius we might be mistaken in thinking he was feeling competitive. In fact, Forest, who had lost his wife and daughter, Amaya, in a terrible car crash, did not want to simulate an Amaya who was not his Amaya. Even if the difference was as small as a single hair on her head. Even though Lyndon’s adjustment made it possible for him to see her again, not just as a blurred figure behind white snow, but in high definition. But eventually, we saw him watch, enraptured, by a completely perfect simulation of a past her, existing in the Devs computer after his number 2, Kate, also implemented Lyndon’s algorithm to the video outputs.
But what does all this have to do with jam and rice pudding?
After finishing Devs, I was thinking about Arcadia and its similarities and differences with Devs. I don’t think Alex Garland was directly inspired by Stoppard’s play, and on the face of it, the play has very little to do with Silicon Valley. But thinking about the two stories together might be fruitful.
Arcadia is set in Sidley Park, a great estate in Derbyshire, and during the course of the play, two apparently separate moments in time take their turn to play out on the stage before eventually overlapping with each other.
In 1809, we see Septimus Hodge tutoring Lady Thomasina Coverley, a teenage mathematical prodigy and the daughter of the Lord and Lady of the house. In the present day, academic Hannah Jarvis is researching a history of the house and its garden. She is using a hermit who lived in the faux-hermitage that was built on the grounds in the early 1800s as the thematic lynchpin for her next book on the Romantic imagination. Her work is interrupted by Bernard Nightingale, a historian working on Lord Byron. Byron was a school friend of Septimus’, and Nightingale believes he may have fought a duel while visiting Sidley Park and killed Ezra Chater, a rather lacklustre poet. Valentine Coverley, the current heir to the house, is using the house’s ‘game books’ (the records of hunting parties’ catches) to crunch numbers for his research into grouse population changes. All three are looking for data to give them insights into the past. But we, the audience, get to see the misunderstandings the academic are falling into as they try to ‘predict’ the past based on the limited data that they come across during the course of the play. For instance, Chater was not killed by Byron. Nightingale, arrogant and sure of his headline-worthy discovery, is blind to the later historical records that mention an ‘E. Chater’ as a botanist because that past doesn’t fit his view of what happened at Sidley Park. In many ways, this group of modern academics are seeing the past in the same static filled way as the Devs team are at the beginning of the series. Back when they are limited to the determinism of one predictable, deterministic, universe.
Arcadia is also a story filled with determinism and despair, just like Devs. In the very first scene, Thomasina is certain that she has had a mathematical revelation:
“If you could stop every atom in its position and direction, and if your mind could comprehend all the actions thus suspended, then if you were really, really good at algebra you could write the formula for all the future; and although nobody can be so clever to do it, the formula must exist just as if one could.” – Thomasina, Arcadia
Of course, Thomasina has only the barest inkling of the kinds of universal difference and analytical engines that were coming. Quantum Computing would also be far beyond the Newtonian physics her calculations were assuming. Squaring the circle of quantum methods and Newtonian predictions in Devs seems to have only been possible through Everett. So again, this leap was certainly beyond Thomasina’s brilliant, but too early, mind.
In this scene, she and Septimus go on to discuss free will, with Thomasina concluding that god must be a Newtonian (delightfully/intentionally misheard by Septimus as “an Etonian”). However, as with the rice pudding and the jam, she’s also realised the big problem of Newtonian determinism. Knowing the ‘tramlines’ (as Forest refers to them in Devs), the direction of travel, does nothing to stop the bad things from happening and they cannot be undone. The jam cannot be unstirred from the rice pudding. Entropy cannot be combated. The heat death of the universe is inevitable. Everything is lost eventually. Even Thomasina herself.
There is an undefeatable fate that also awaits her ahead on the tramlines. We learn from Hannah and Valentine in the present day that Lady Thomasina Coverley died in a fire the night before her seventeenth birthday. Just moments later, we watch her dance with Septimus on that fateful night, as Hannah and Valentine discuss the mysterious figure of the hermit who hid himself away in the fake-ancient hermitage in the gardens. The Genius of the Place, as Hannah’s book will eventually be called, using both meanings of the expression. The genius of the place was a man who spent the rest of his life attempting to disprove the second law of dynamics that Thomasina had stumbled upon forty years early, in order to return back down the path he had already trodden. To remove the jam from the rice pudding. And perhaps, like Forest, to bring back a lost one:
Septimus, as the hermit Paulus, hoped to use good English algebra to turn back time. Septimus was unsuccessful. Thomasina was gone forever.
However… the play offers us a simulation of the past. Thomasina dances again as the past is replayed simultaneously with the present inhabited by the academics. The play also explores Chaos Theory (more prominent in the 1990s and replaced in favour by more quantum theories now), and the way in which the objects from the past (important clues and pieces of data) are left on the table by the 1809 characters evokes increasing entropy and chaos.
Forest, as the head of Amaya, hopes to use good Quantum Computing (not American Quantum Computing, he’s quite clear he sees nations and states as hindrances to his work) to turn back time. Forest is successful. Amaya is back (forever?).
The series ends after the death of Forest and Lily and their resurrection in the Simulated Universe Forest has created inside Devs/Deus. Up until this point, I haven’t discussed Lily’s role very much. She is the main protagonist, and her actions lead to a point of uncertainty that the algorithms can’t see beyond; a point with no. That point turns out to be a moment of actual free will, as Lily refuses to kill Forest as was predicted by Devs. Even so, she and Forest still die and are then resurrected in the Simulated Universe. For all the uncertainty Lily introduces to the deterministic system, she still proves Forest’s personal prediction right. She tells him, “The problem with people who run tech companies is that they’re fanatics. They end up thinking they’re messiahs”. He responds to this in the Devs simulated scenario of their deaths, first telling her that Devs is actually Deus, and that the thing about Messiahs is that they get resurrected.
Of course, since Lyndon, the system has been ‘infected’ with multiverse theory. So, in effect, Lily’s decision not to kill Forest spins off a whole new universe. Hence Devs not being able to see beyond that point of choice. However, this new universe is a Simulated Universe, inside Devs itself. It’s a nice one, as Forest points out to the confused Lily newly discovering their world. Its more heaven than hell, certainly. Amaya and Forest’s wife are back. As are other characters who died during the series. It is a paradise.
Or we might see the new universe not as paradise or heaven, but as the world beyond the Garden of Eden. Kate refers to Lily as having committed the ‘Original Sin’ in making her choice not to kill Forest. By which she meant that Lily had acted in a moment of complete free will, just as Adam and Eve did in the Garden in the Old Testament. However, instead of being ejected from the Garden of Eden and ending up somewhere worse, Lily and Forest are resurrected after her choice in somewhere much more utopian.
However, Et in Acadian Ego. “Even in Arcadia, there am I”.
This is the title to a 1637 painting by Nicolas Poussin. It shows four shepherds in a beautiful, idealised, pastoral scene. They are examining a tomb: a reminder of death, even in life. Even arcadia – paradise, utopia, etc – contains death. Initially, Tom Stoppard named his play Et in Arcadia Ego, before shortening it to just Arcadia. The entropy and chaos that Thomasina predicts and that Septimus spends his life trying to refute is also in the paradise of Sidley Park and its romantically renovated gardens, hermitage and all. Thomasina’s death proves that. But is death also in the Simulated Arcadia that Forest and Lily end up in?
One strangeness in the series Devs is that Lily, a computer programmer based in Silicon Valley, seems to have no awareness of Simulation Theory or computational theories of the universe (or even determinism at one point). It was a little like those moments in Zombie movies when the main characters are faced with a shambling undead being that bites humans and is susceptible to a headshot but can’t bring themselves to say the ‘Z’ word. As the audience’s stand-in, Lily might be forgiven for this, we need a fair bit of exposition to follow the plot. But Forest really should have known that the universe he was simulating inside Devs might be open to entropy and death. It would be there either through the breakdown of the physical set up in the ‘base reality’ of the original Devs building and Amaya as an organisation. Or, entropy might leak in through the original data itself, as ‘Et in Terra ego’. And of course, the Simulated Universe could not be entered without Forest and Lily dying in the first place. Another programmer character Stewart, recites Philip Larkin’s poem Aubade (mistaken by Kate for Shakespeare, and proving for Stewart that the people creating the future know nothing about the past), a poem also about the inevitability of death. Even religion, “That vast moth-eaten musical brocade created to pretend we never die” (Larkin), cannot help us to escape that inevitable. Even Devs, a new religion writ in golden temple and halo-ed trees and complete with its own Messiah, can’t either.* And it is notable that it is Stewart who kills Lily and Forest.
But there is also the argument that mind-uploading (in this case, even the most perfect mapping of all the possible data of a person) and virtual resurrection could never recreate the same person anyway. So even if Forest and Lily had not died, they died the very moment they were uploaded.
The religious narratives were ever-present in Devs, and very cleverly explored. Forest and Devs were both compared to gods/messiahs, and Lily was compared to Adam and Eve in making the first-ever completely free choice. Arguably, that moment of choice was more divine than human, as Lily’s choice led to the creation of an entirely new universe.
But then, if Everett was right, we’re all gods making new universes, with every choice we make.
* thanks to @huwcdavies of the Oxford Internet Institute for discussing this blog post with me and making me think further on the significance of the Larkin poem 🙂
This is a speech I gave last night, as the Junior Research Fellow in Artificial Intelligence, at a Homerton dinner themed around the Roaring ’20s in celebration of entering the 2020s.
Welcome to the Roaring ’20s.
There have been approximately 21 decades we could call the ’20’s – depending on whether you think Popes are worth listening to or not on matters of timekeeping. But only one ’20s is commonly called ‘roaring’ for its exuberance and joi de vivre.
On this VERY day in 1920 the last meeting of the Paris Peace Conference took place, and while negotiations and treaties continued to be put in place over the course of the next decade, for many the 1920s – the roaring 20s – kicked off with this moment of hopeful restitution after what many came to think of as the first-ever ‘world war’.
Coming fast on the heels of 1920 was a decade of growth and prosperity. Drinks were downed – even during the Prohibition in America – and hemlines shot up.
Deferred spending led to a boom in construction, a boom in credit, and a boom in consumer goods. Electricity, cars, movies, planes… everyone moving faster and faster, and not just while dancing the Charleston.
After the sadness of World War 1, Jazz burst on the scene to break the apparent rules of music and to bring together people of different ethnicities and sexualities in underground clubs. Authors wrote about fabulous parties and pretended that they’d gone to them. Quantum physics turned up to make everything weird before it would cut down the cat population in 1935. Or not. Bright Young Things dropped out of Cambridge (and the Other Place) – possibly after they had popularised the ‘bring your own bottle party’. And women’s Suffrage advanced – for some.
But, the ‘roaring’ was eventually quietened.
In 1929, the flamboyance and frivolity came to a crashing end as a bubble of excessive speculation and complex world-wide debt burst violently, sending everyone into a new decade characterised by financial and emotional depression.
Here, at the beginning of a new ’20s, we need to think about whether this decade should roar like its 100-year-old namesake, or whether we should learn from the excesses and acceleration of previous Bright Young Things to tread more warily through the next ten years.
Now, I assume that I have been invited to speak to you today because artificial intelligence is one of those things that we have been expecting to be a part of ‘THE FUTURE’ for a long while. Along with flying cars, jetpacks, and our alien neighbours, AI is another not-quite-here-yet treasure, or terror, that could just be around the corner, in the world of tomorrow.
Throughout the 20th century, writers have imagined what the wonderful world of the 21st century would look like. To many people, any year starting with ‘20’ appeared as inherently fantastical and futuristic. It certainly did to me as well when I was younger. The year 2000 was ‘the future’ in more than just the chronological sense.
I’m assuming that most of you weren’t actually born in time to welcome in the new millennium on 31st December 1999. So, very few of you could have genuinely partied like its ‘1999’ as Prince first told us to in 1982.
I, however, was. In fact, at the time of that great threshold, I was merely one term into my first year at Cambridge.
I travelled home after Michaelmas term and re-joined my school friends back in Portsmouth. I drank copious amounts of peach schnapps and Lemonade at a friend’s house party before dashing outside during the bongs to watch a new millennium come into being – via fireworks of course – over the harbour. I was nineteen and 100% certain that the new millennium held for me many exciting things. Wealth and fame as a Hollywood screenwriter. A large and expensive flat in Soho. Another in New York, of course, with the mansion in LA. There would be Oscars and other assorted awards, which I would receive with suitable modesty and thanks to the little people.
And I had high hopes for the long termness of my relationship with a third-year I’d started dating during fresher’s week!
None of those things happened! But I’m not bitter. Not really.
In fact, I probably should have paid more attention to the science fiction writers who told me back then in the books I loved that my future in 20-something was going to be full of robots.
Because my ‘now’ certainly is. As the Junior Research Fellow in AI here at Homerton, I think about robots and AI much more than I could have ever predicted I would do, way back in 1999.
But also, because the world is genuinely full of robots.
They aren’t exactly walking down the street, helping old women (like myself), to cross the road, but they’re around.
A hundred years ago, at the start of the roaring 20s, Karel Capek gave us a name for such artificial beings in his play, “Rossum’s Universal Robots”: ‘robota’, from the Czech word for ‘serf’. He was struck by the fast growth and increasing dominance of the ‘factory’ and extrapolated from the treatment of the humans in such intense work environments to an imaginary being that could rise up against its overlords.
Capek’s ‘robota’ was made from a grey synthetic flesh, but by 1926 Edmond Hamilton wrote The Metal Giants, in which a computer brain running on atomic power creates an army of 300-foot-tall metallic robots. Such science fiction accounts have given us an impression of what the robots of 2000 and something would be like.
First, they’ll be embodied, as Capek’s robota were. But after the 1920s we came up with other robotic embodiments. There’ll be sexy ones, strong ones, fast ones… but most of all there’ll be ones that are hard to distinguish from humans.
Simon Schaffer has previously drawn a connection between Alan Turing’s famous test for computer intelligence with Cold War fears of secret agents ‘passing’ for ‘us’. In the 1980s Philip K Dick wrote us a new Turing test in his novel Do Androids Dream of Electric Sheep. The Voight-Kampff empathy test was necessary as the replicants – again bioengineered like the ‘robota’ of R.U.R. – could pass unknown among us and need to be detected before they can hurt us.
Second, in having all these attributes, the robots might just be, well… better than us. And better is dangerous. Capek wrote the first popular robot factory rebellion and in 2019 Terminator: Dark Fate encouraged its audiences to draw a causal line between the factory worker who loses his job to a much more efficient automaton and the latest iteration of the deadly Skynet, Legion.
Of course, the factory robot isn’t embodied in the same way as the Terminator. And this is where 20th Century science fiction has failed us.
But then, science fiction is never about accurate prediction, but instead, it serves as a commentary on where we are now. Karel Capek wrote about the factory worker made of synthetic goop to focus us on industrialisation and dehumanisation.
However, Capek, and others, have inadvertently helped us towards a certain kind of face-blindness when it comes to robots and AI.
A robot-blindness, perhaps?
Almost every single one of you uses a robot at least once a week. Or… at least I hope that you do!
A washing machine meets the basic requirements of a robot. A washing machine has settings that allow you to alter its automatic programming to perform a series of procedures. And yet, I think, unless your washing machine also spoke to you, folded your clothes and put them away, and then asked you if you’d like a nice martini like a proper 1920s Jeeves, you wouldn’t think of it as a ‘real’ robot. This would require a few more bells and whistles, including genuine ‘AI’.
Conversely, though, when we do see an anthropomorphised robot – with two arms, and two legs, and a flubbery face – we are more inclined to see it as having that mysterious thing called ‘intelligence’.
Take Sophia, the Hanson Robot. She is by no means the most advanced form of AI available. Her interviews are scripted, her appearances highly orchestrated, and her social media account run by a human. But because she fits so neatly into our sci-fi informed expectations of a 2000 and something robot, we start to think that perhaps the future promised by those films and books set in 2020 is actually here, now.
But that is another form of robot-blindness and one which distracts us from all the examples of AI that aren’t shaped like humans.
And the various AI implementations that are insidious in our systems and our processes in the 2020s are the product of an accelerationist view, much as the boom in other technologies was in the 1920s.
You might have come across the expression, “Move fast and break things” – it was an internal motto for Facebook. Well, as we enter the 2020s we might see that some pretty important things are being broken by moving so fast. The roaring 1920s rushed millions into debt through speculation and commodification, and movements begun to help the people evolved in dangerous, popularist, and nationalist directions, leading to a second World War.
And the roaring 2020s are growing out of a decade where already social media and its algorithms have been weaponised by familiar ideologies.
A decade where decisions about our capability to do a job, to take a loan, or even to access information online were already being automated by non-transparent systems. A decade where unseen influence pushed and pulled us towards particular political candidates. A decade where AI was already being trusted before it was even as remotely as intelligent as the human next to you because it came packaged as the next Bright Young Thing and appeared on the Jimmy Fallon show.
So, will 2020 be a roaring decade of prosperity and glamour? Or will the excesses of accelerationism and technological change dash us, Charlestoning all the while, into a brand new Great Depression?
I want to suggest that there can be more than one way to roar.
Perhaps, instead of just having excessive parties – or bread and circuses – we will roar against the changes that would reduce us to parts in a greater machine. But I say this without advocating extreme Puritanism and a new Prohibition.
After all, the roaring 20s also involved activism, charity, and progress for many underrepresented groups. Although, there was also uneven and slow progress in many cases.
But if we are to try to imagine, as science fiction authors have done, a future containing robots, we should at least endeavour to make sure that we are not the cold, mechanical, beings that they wrote about. We should recognise the robotic around us, and within us.
In the roaring 2020s, it might seem like the future is here and now.