In Roald Dahl’s short story ‘The Great Automatic Grammatizator’, an inventor named Adolph Knipe constructs a novel-writing machine. Knipe’s fantastical and perhaps excessively gothic contraption can be manipulated by various pedals, levers and organ stops to produce detective stories, historical fiction, Westerns, tales of the sea and so on – in the style of Hemingway, Faulkner, Joyce or whoever you like. The Grammatizator’s owners plot to supplant the world’s novelists – ‘squeeze ’em out’, ‘exactly like Rockefeller did with his oil companies’, one of them enthuses. This being a Dahl story – and therefore dark and cynical – the plan is a virtually unqualified success. Dahl writes that in ‘the first full year of the machine’s operation it was estimated that at least one half of all the novels and stories published in the English language were produced by Adolph Knipe upon the Great Automatic Grammatizator’.

The Grammatizator is not quite here. But new generative artificial intelligence technologies, such as ChatGPT, are capable of producing plausible essays and pastiches of poems. Writers are understandably alarmed. In Hollywood, screenwriters have gone on strike against studio plans to use large language models to write television scripts. Upsettingly, for those in my profession, AI already writes competent (if not always accurate) journalism. Some experts predict that a machine will be capable of writing a bestselling novel within decades or years.

If art and literature matter to you, no thought could be bleaker. To me, the prospect of a machine capable of writing a book seems almost unbearably sad. An assault on a profound aspect of what it is to be a human being. But need we be as pessimistic as Dahl? A machine may write books, but will it produce the higher form of writing we call literature? I am sceptical. For to believe a machine is capable of producing literature is to misunderstand what literature is and why we love it.

Let us take, for instance, this passage from Matthew Arnold’s great poem ‘Dover Beach’:

Ah, love, let us be true

To one another! for the world, which seems

To lie before us like a land of dreams,

So various, so beautiful, so new,

Hath really neither joy, nor love, nor light,

Nor certitude, nor peace, nor help for pain;

And we are here as on a darkling plain

Swept with confused alarms of struggle and flight,

Where ignorant armies clash by night.

The profoundly moving poem records a familiar human sentiment. But no machine with access to the whole corpus of human writing should find it difficult to construct a description of despair – the subject crops up with melancholy frequency. And true, the final image is especially skilful. But we have no reason to believe literary skill will be permanently beyond the abilities of artificial intelligence already capable of writing a sonnet about spinach in the style of William Shakespeare.

But the poem is also a record of the noble, melancholy soul of Matthew Arnold. It matters as a human artefact, and an appreciation of Arnold’s lines is inextricable from an appreciation of the human personality that created them. Even the reader’s thrill at the poem’s technical virtuosity is inextricable from the human intelligence that conceived it. You could give me a hundred years and I could not write the final lines.

One’s appreciation of a work of art is often accompanied by the thought: ‘Wow, how did they do that?’ Part of the excitement of reading poetry or looking at a painting is the thrill of watching a human talent pushed to its utmost. The feeling is not dissimilar from the way an appreciation of athletics derives from our interest in what the human body can be made to do. Indeed, at the ancient Olympic Games there were prizes for poetry as well as sport. It is true that you could write a poem using artificial intelligence. Equally, the competitors in a 100-metre race could augment their physical skill by getting in a car and driving to the finish line – but that would be a different race. The existence of Formula One has not diminished the appeal of the foot race sprint.

There is an analogy here with chess. The game (or perhaps even the art) of chess was ‘disrupted’ by artificial intelligence almost three decades ago. Modern AI programmes make the best chess players look like pathetic amateurs by comparison, but this has done little to diminish the appeal of the game. More people play than ever before. And few chess players in history have achieved the fame of the greatest living player, Magnus Carlsen. Online, millions have watched Carlsen’s greatest matches. Games between computers tend to attract much smaller audiences of dedicated nerds. Many more people have heard of Carlsen than have heard of the chess AI AlphaZero, but Carlsen is vastly the inferior player.

Human beings have a profound bias towards other humans. As the psychologist Steven Pinker puts it, ‘we all have deep intuitions about causal connections to people’. It is for this reason, Pinker says, that collectors will pay a huge premium for JFK’s golf clubs, for example, even if they are identical to those owned by a less well-known player. It turns out that chess fans are not purely (or even primarily) interested in the perfection of a game; rather, they are interested in the potential of the human mind and the drama of personality. Examples of this bias are everywhere.

Any listener to a public phone-in show on the radio knows the callers are not principally interesting as analysts of current affairs. Their opinions fascinate us because they are the views of human beings. Even an AI so powerful that it solved the question of immigration or Brexit would be unlikely to abolish the radio phone-in show.

We may justly wonder: how interesting is intelligence to human beings? Interesting enough, certainly, but rarely the most fascinating thing. It is the same with art. Indeed, few people are as intriguing to us as famous singers, actors and writers. Newspapers are filled with gossip, sex scandals and speculation about the private lives of celebrities. Our fascination with art is deeply connected to our fascination with the human personality.

These personalities matter to us because art is profoundly human – as most great artists have understood. Philip Larkin once said: ‘I write poems to preserve things I have seen/thought/felt (if I may so indicate a composite and complex experience).’ Art, he wrote, is ‘permanent communication’. In his treatise of 1897, What is Art?, Tolstoy reaches a similar conclusion: ‘Art begins when a man, with the purpose of communicating to other people a feeling he once experienced, calls it up again within himself and expresses it by certain external signs.’ If a work of art does not preserve a human experience, is it really a work of art at all?

‘Dover Beach’ is a consoling, not a depressing, poem because it preserves a human experience. I am comforted by it because it is a verbal artefact that has been skilfully constructed to connect me to another person. Alan Bennett puts it perfectly in his play The History Boys: ‘The best moments in reading are when you come across something – a thought, a feeling, a way of looking at things – which you had thought special and particular to you. Now here it is, set down by someone else, a person you have never met, someone even who is long dead. And it is as if a hand has come out and taken yours.’ Tennyson put it like this:

So word by word, and line by line,

The dead man touch’d me from the past,

And all at once it seem’d at last

The living soul was flash’d on mine

A feeling reconstructed by a machine cannot console in this way because it has no human hand to reach out to us. Perhaps even a conscious machine could not really make art, because its feelings would not be human ones. I can imagine a machine that could entertain. I cannot imagine one that could console.

This instinct for the human is not merely a philosophical abstraction. It is one of our most instinctive responses to art. Watching a YouTube video of the Mozart opera Così fan tutte recently, I was struck by the most up-voted comment: ‘Absolutely moves me to tears every time I hear it. Whenever you see humans doing terrible things to each other or themselves I listen to this and realise what humans ARE capable of. People are making beautiful noises!’

I suppose one can imagine ways we might be duped into this response by means of deepfakes and so on. But the deception would have to be elaborate, and perhaps involve the faking of a human author along with their characteristic signs of existence – interviews with newspapers and so on. And if no legislation effectively compels publishers to state which content is made by humans and which by machine, we will have more urgent problems on our hands than the meaning of literature.

It is curious that machine art should have arrived at a time when our cultural obsession with the human personality in art is more powerful than ever. The writer David Shields’ useful phrase ‘reality hunger’ describes the 21st-century yearning for the authentic and the real.

As sales of literary fiction decline, the memoir flourishes. As soap opera viewership slumps, reality television booms. Even the novel itself has succumbed to this trend. Writers like Karl Ove Knausgård, Ben Lerner and Rachel Cusk have pioneered the new form of autofiction – novels that acquire their fascination and authority from their connection to life. The power of the novels that make up Knausgård’s My Struggle series derives from the reader’s suspicion that everything, or almost everything, in the book is real. Knausgård’s genius for artistic banality – the long passages describing, say, the making of a cup of instant coffee, or a journey through the Swedish countryside with fractious children, rests on our belief in the authenticity of these moments. Knausgård could have made up more dramatic incidents, but boring reality is more interesting to his readers. Such is human nature.

Similarly, many reality television shows are significantly less eventful and artful than their fictional rivals. And yet, in the early 2000s, Big Brother enthralled the British public with its endless broadcast of people in a house not doing a great deal. What mattered was that these were real human beings.

In a future flooded with machine-produced content we may place a higher value on human art than ever before. In his classic essay ‘The Work of Art in the Age of Mechanical Reproduction’, Walter Benjamin speculated that the advent of machine reproduction of images would damage the almost magical aura of authenticity that surrounded original artworks. What is the point of Van Gogh’s Sunflowers if you can buy a thousand cheap prints of it for a fraction of the price? Benjamin, of course, was wrong.

Precisely the opposite happened. The proliferation of cheap prints increased the mystique of authentic artworks and caused their market value to rocket. In the latter half of the 20th century art became obscenely expensive – its value deriving from its contact with the original human genius. Perhaps the advent of AI art will have a similar effect. A culture flooded with machine-produced content may be one that places an extraordinary premium on the human.

The future of many careers may turn out to depend on the value we place on that illusory, illogical human factor. Could we trust the justice dispensed by a robot judge, even if we were told it was measurably more reliable and less biased than that of a human? Will the opinions of a robot newspaper columnist matter? I sincerely hope not. Most of all, I hope we decide that human art still matters.

A world of machine art would be an eerie one. Art connects us to one another. We cannot, and we should not, replace that connection with an uncanny simulacrum of it. For without the art and literature of human beings we are left in the situation described in another great Matthew Arnold poem, ‘To Marguerite: Continued’:

Yes! in the sea of life enisled,

With echoing straits between us thrown,

Dotting the shoreless watery wild,

We mortal millions live alone.

Art and literature are the bridges between those human islands. Let us hope we keep them in good repair.

If you enjoyed this essay by James, listen in through the link below to him in conversation with the EI team: