SUNDAY, 8 SEPTEMBER 2024

This week a picture came up on my Facebook feed which called itself a painting in the style of Edward Hopper. My immediate reaction was that it looked like Hopper but didn’t feel in the slightest like Hopper. The woman in the railway carriage was much too calm and composed; there was none of that unease and loneliness of a classic Hopper, nor any of its sense of inner emptiness. I suspect many other posters felt the same as me because several of them openly wondered if this was an AI-generated work.

There used to be a saying that ‘the camera doesn’t lie’. This wasn’t always true, of course, even in the days before photoshopping – careful cropping, for example, could create a variety of different realities, or at least interpretations, from the same photographic negative. But in general people believed a photo: it offered a technological version of ‘I saw it with my own eyes’, and it was therefore assumed to present the truth.

Few of us are so trusting now. We know that not only photographs, but even whole videos, can be concocted, and that if we see a film of Kamala Harris beating a cute puppy with a stick, we may well be looking at a fake. Unable any longer to believe our own eyes, we have come to rely on experts who declare whether a picture is real or has been tampered with, or if a video is fabricated. The problem with this is that we no longer have responsibility for what we see and what we therefore believe to be true; we need to pass that responsibility on to a third party. And then we are trapped in an infinite regress: how do we know that these experts are real and can be trusted? Who, or what, monitors the monitors?

In the world of art, this has updated and exacerbated the problems of detecting a forgery. At the moment, perhaps, like some of the people on my Facebook post, we may feel able to sense the difference between an AI-generated Hopper and the real thing. But as AI develops, who is to say that it will not be able to reproduce not merely the surface of a work of art, but also its essence? And while works by historical figures like Hopper may be relatively easy to verify or dismiss because we can demand very solid evidence to show that a newly discovered painting is genuinely his, living artists and writers are unlikely to be vetted so rigorously.

AI certainly complicates the issues of copyright and intellectual property. In a world where anyone can task AI with writing a poem or painting a picture, how can we be confident that a new work is created by a human being? It can be argued that AI is currently only churning out the kind of dross that poor genre writers have been doing for decades － boy meets girl, boy loses girl, boy and girl get back together and walk off into the sunset. I’m sure AI will be able to crank out cheesy romantic novels and scripts for action movies and good-versus-evil fantasies. But how will human artists earn a living in this world? And how do we prevent the further decline of art and literature into banality, cash-cow sequels created by dull middle-managers in marketing departments who have the aesthetic sensitivity of a friendly dictator and the creative spark of a concrete mixer?

In painting, at least, there is the requirement for a physical object to be created. But eventually AI will be attached to robot arms or a 3D printer, I’m sure, and even this role for the human being will be lost (although the physical nature of the art work would allow for more opportunities to check its veracity through processes such as material appraisal of the canvas). Literature, especially fiction, is under much greater immediate threat. On my Facebook feed, I regularly see adverts inviting people to ‘write’ and publish fiction by using AI (and I don’t mean using it as a guide or a source of inspiration, but literally giving AI a set of instructions, pushing a computer key, and out pops a novel). Why pay writers for the script of the latest dreary sequel if the sausages can be squeezed out just as efficiently by AI?

There has always been a delicate balance between reality and fantasy in Art. Part of our enjoyment of literature or theatre or film comes from entering a world which we know is make-believe. This becomes explicit at times: Calderón’s Life is a Dream, The Matrix, the twists and turns of Borges, the trompe d’oeil of Op Art, the work of Magritte, the impossibilities of Escher. But our pleasure always rests on a conviction that there is a real world to go back to, a real world we are taking a break from. Like the big dipper at the fair, we enjoy its otherness because we feel safe. We are not going to wake up to find ourselves transformed into a giant cockroach.

But this sense that AI is undermining the solid foundations of our lives is something that is affecting more than Art. There have always been individuals who questioned the reality of reality － Zhuangzi and his butterfly dreaming it was a man － but they have tended to be marginalised on the hermetic fringes of society. Their ideas had little traction with the vast majority of people who treated these ideas － if they thought about them at all － as conceits, the odd twitterings of mystics and madmen. But soon we may all be living in a world where the ground under our feet is unstable and we have very little confidence about what is real and what isn’t. Already so much of the world we inhabit is virtual (walk down the street and see the pedestrians with their noses pushed up against their mobile phones), and we often spend more than half of our waking lives staring into a screen and communicating with people we have never met in person. Slowly we are losing the security of feeling that there is a normal reality that we can return to when virtuality is over. It is never over. This radical uncertainty is our new reality.

I haven’t even expanded on the political threat of AI in a world in which life is increasingly mediated. Our leaders and politicians are now figures we see on TV or on the internet, and this is our only way of judging them. Again this makes us dependent on experts who can separate the wheat of the truthful from the chaff of the liars, but who pays these experts and who controls them? Will they really be disinterested? Will we be able to trust this media class to act as honest go-betweens? Or will he who pays the piper call the tune? I recognise that this gatekeeping role has always existed, but the gatekeepers were at least visible and, in theory at least, answerable to us. In a world of the huge online conglomerates that shape our daily lives, I am not confident that this is any longer true.

And what will this do to us psychologically? Eliot’s dictum that human beings cannot bear too much reality is a favourite quote among pessimists like me. But perhaps it is equally true that  human beings cannot bear too much unreality. Will we descend into madness in a world where we can never be certain that anything is real? Are Deleuze and Guattari, for all their impenetrable language and pseudo-scientific pretensions, or earlier counter-cultural figures like Laing and Szasz, correct when they argue that we are already on our way to a schizoid world where we have lost all our moorings and only the madman is sane? Personally I find it increasingly difficult not to agree with them that our modern, mediated world is fundamentally sick.

Perhaps civilisation will not end in the bang of an asteroid hitting the earth or a thousand nuclear warheads, but in the whimper of a billion AI mutations which undermine and eventually shatter our sense of reality. Whom the gods would destroy, they first make mad.