We need to talk about AI.
I say ‘need’ rather than ‘want’, because if I’m honest, a big part of me would rather not. I am not an enthusiast for AI. If it were up to me, on balance, I would rather it didn’t exist.
But we’re now at the stage where AI is unavoidable. Big changes are already here. And they are being felt particularly keenly in the arts and creative industries.
Some of my friends have been fired up for years about the creative potential of AI. It’s always good to see a friend full of enthusiasm, but listening to them, I didn’t feel it myself.
I’ve been wrestling with AI for over a decade — my earliest concerns were the big existential ones. But for most of us, the urgent problem isn’t superintelligence. It’s what’s already happening to creative work.
Why creatives have good reason to hate AI
Let’s start with copyright, which is currently a large bone of contention, playing out in debates, protests and lawsuits around the world.
Here in the UK people are getting very exercised about the issue of creators’ work being used to train AI models without permission – whether or not this constitutes copyright infringement, and if so, whether anyone is going to do anything about it.
Creatives are protesting. Furious debates have been raging in both Houses of Parliament. Open letters have been written and a petition has gathered the signatures of Kazuo Ishiguro, Simon Rattle, Kate Bush, Stephen Fry, Alan Ayckbourn, Antony Gormley, Roger McGough, Robert Smith and other stars in our creative firmament.
This is the point where the debate stopped being theoretical for me. When I saw an article in The Atlantic alleging that Meta had trained its AI models on the LibGen database of pirated ebooks, I typed my name into the helpful search box on the Atlantic website, and saw that my four books for creatives were in the database.
It took me several years to write those books, and many more years to gain the experience on which they are based. I am the publisher as well as the writer, which means I also spent thousands of pounds on editing, proofreading and design costs.
Call me old-fashioned, but if a multi-billion-dollar corporation wanted to take my books and use them – as is alleged in the Meta case – it would have been nice of them to ask.
Whether or not Meta did this, I wouldn’t be terribly surprised to discover that other companies have done something similar.
So I have skin in the game. If you are reading this and feeling angry about your work being used without your permission, and fearful about the consequences of AI for you as a creative professional, then I have had a taste of how you feel.
How has the UK government responded to this kind of allegation against the AI industry? It has made a concerted effort to weaken UK copyright law, to create a special exemption allowing AI companies to use creators’ work to train their models, without asking permission or even telling creatives their work has been used in this way.
And of course, it is these very models that are now being trumpeted as the ‘future of creativity’, and causing millions of creatives to fear for their livelihoods.
There are newspaper stories about creatives being replaced by AI. Some researchers are finding evidence of a decline in creative freelance work since the launch of ChatGPT in 2022.
It feels like a triple whammy:
Theft: Taking our work without permission to train their models.
Threat: Using those models to develop machines that threaten to replace us.
Insult: Flooding the culture with hideous AI slop.
I’ve signed the petition. And I would love the debates and petitions to have some kind of effect.
But I’m not holding my breath. Looking around the world, it’s clear where the balance of power lies between the tech and creative sectors. We can’t depend on governments or courts protecting our rights.
No one is coming to save us.
Why we can’t ignore AI
Even if we loathe AI — even if we refuse to use it on principle — it is already reshaping the conditions of creative work.
Clients and employers are adopting it, and expectations are shifting with them: faster turnarounds, more variations, more “Can you just…?”
At the same time, AI is being baked into the tools we already use. Writing apps, design software, search engines, email, editing suites — features are arriving by default, often switched on before we’ve even decided what we think about them.
And whether we like it or not, other creatives are using it. Not because they’re better artists, but because it can accelerate the dull bits — research, rough drafts, mockups, options — and that changes the market we’re working in.
Meanwhile the culture is being flooded with slop: more content, more images, more noise. That puts downward pressure on commodity work and raises the premium on taste, trust, and a recognisable human voice.
So yes: we can refuse to engage. But we can’t opt out of the consequences.
The AI genie is not going back in the bottle. There is no turning back to the old world. We need to face this with our eyes open.
There is no world where humans don’t create AI
It’s natural to be preoccupied with the current state of the AI industry, and the consequences for our lives and livelihoods. And to wish for a world without AI – or at least with wiser, more ethical and more trustworthy people leading the industry. But that is not the world we are living in.
If we look back into the past, from the earliest times shrouded in myth, to the long sweep of human history and the progress of technology, the development of AI begins to look inevitable.
Back in the Garden of Eden, Adam and Eve could have been perfectly happy, munching away at the fruit of the Tree of Life, forever. But they couldn’t resist the one thing they were forbidden by God – the fruit of the Tree of Knowledge.
And once they had eaten the fruit, their eyes were open. They saw that they were naked. They were expelled from Eden, to live by the sweat of their brow and the ingenuity of their brains.
Generations later, humans had gone from hunting and gathering to planting and sowing, building city-states and fighting over them. In the first millennium BC, when metalsmiths were leading technologists, hammering out weapons and armour and ploughshares and other tools, humans were already dreaming of artificial intelligence.
In Book 18 of Homer’s Iliad, around the 8th century BC, there is a description of gold automata (Homer’s word) ‘filled with minds and wisdom’ that sound remarkably like intelligent robots, serving Hephaestus, the Greek god of blacksmiths and other artisans:
Handmaids of gold attending him, resembling in all worth
Living young damsels, filled with minds and wisdom, and were trained
In all immortal ministry, virtue and voice contained,
And moved with voluntary powers; and these still waited on
Their fiery sovereign,(Translation: George Chapman)
So creating artificial life is one of our oldest dreams. It’s such a fascinating problem, it attracts bright minds like moths to a flame. And not just writers of poetry and fiction, but inventors trying to do it for real.
Around 400 BC, Archytas of Tarentum is reputed to have made a steam-propelled wooden bird that could fly about 200 metres. Just over a century of later, Ctesibius built the first known feedback-control device – a self-regulating water clock. And by the first century AD Hero of Alexandria is said to have created a programmable theatre show, featuring moving statues, singing birds and automatic temple doors driven by water, steam and weighted ropes.
It makes Mary Shelley’s Frankenstein (1818) and Charles Babbage’s Analytical Engine (1837) look positively recent.
Since we built the first machines, we have dreamed of creating thinking machines. It’s as if we are hard-wired to do so.
Our progress towards AI has followed a very long and drawn out exponential curve, measured first in millennia, then centuries, then decades – to where we find ourselves today, when things have accelerated alarmingly since the launch of ChatGPT in 2022.
The rate of change is speeding up, and is projected by many experts to accelerate even further, to the point where we reach Artificial General Intelligence (AGI) – with capabilities comparable to humans on any cognitive task – and beyond that, the fabled singularity.
And whatever we think of some of the people developing the technology, the technology itself feels inevitable.
As inevitable as microbes appearing in the sea millions of years ago.
As inevitable as organisms growing eyes and gills and fins and arms and legs.
As inevitable as the first fish venturing onto the land.
As inevitable as the hand reaching out for the forbidden fruit. Because human beings will always eat the apple.
There is no world in which we didn’t eat the apple. No world in which we didn’t evolve to become intelligent and conscious and insatiably curious to find out more and more about our world and what we can create.
No world in which we don’t want to know the answer to the question: can we build something more intelligent than us?
So from this perspective, AI as a fundamental technology is neither good nor bad. AI itself is neutral.
If we ignore AI, or demonise it because of the people involved, or specific ways it is being developed and applied, or how the politics and economics are playing out in our world, we are not seeing it clearly. We risk missing the real threats, and the real opportunities it presents us with.
And if we are tempted to put our heads in the sand, rest assured the tech bros and politicians are not. They are setting the terms of the brave new world, whether we like it or not.
If we as artists and creatives can’t take our bearings from them, who can we look to?
What would Bowie do?
When I find myself agonising over a big decision, to do with my artistic work or my career, I have a question I use as a thought experiment:
What would Bowie do?
I’ve long admired David Bowie as an artist, for many reasons, including the fact that he was always leaning forward and looking into the future and doing his best to channel its signals for the rest of us.
He was singing about the downside of space travel in 1969 and the threat from AI in 1970. His big breakthrough came in 1972, when he appeared on UK TV sets as a gender-fluid Starman for three and a half minutes that changed the course of music.
And he didn’t just sing about the future – he continually experimented with new techniques and technology, to disrupt his thinking and spark his imagination.
In the seventies he was using William Burroughs’ ‘cut up’ technique and Brian Eno’s ‘Oblique Strategies’ cards.
In 1976 he left Los Angeles for Berlin, and embarked on a trilogy of albums using techniques from the emerging electronic and ambient music scenes.
By 1994 he was using a program called the Verbasizer to randomise his lyrics and blend them with other texts.
That same year he released a CD-ROM of the video to his single ‘Jump they Say’, encouraging fans to remix it, with the instructions: ‘Sculpt the album of the future… Don’t be passive; be interactive.’
The single was from 1. Outside, a dark futuristic thriller in album form, where he played a series of characters, using computerised vocal distortion on the record and visually distorted photos on the artwork.
In 1996 he was the first major artist to release a single, ‘Telling Lies’, for download online. (I tried to download it, but it took forever on a 56K modem, so I gave up and bought the CD, but I appreciated the gesture.)
In 1999 he gave a remarkably prescient interview to Jeremy Paxman, where he shattered Paxman’s assumption that the internet was ‘just a different delivery system’:
‘I don’t think we’ve even seen the tip of the iceberg. I think the potential of what the Internet is going to do to society, both good and bad, is unimaginable. I think we’re actually on the cusp of something exhilarating and terrifying… It’s an alien life-form… it is going to crush our ideas of what mediums are all about.’
So how do I think Bowie would respond to the creative challenge of AI?
I think he’d be all over it.
He wouldn’t be naive. His 1970 album The Man Who Sold the World included ‘Saviour Machine’, a song about ‘President Joe’ who hands over power to a super computer that is supposed to make life easy for human beings – but it turns on them in disgust and threatens kill them all.
So I’m pretty sure he’d be alert to the risks of AI.
Maybe he would add his name to the current petition, alongside the other stars, and speak up for a better deal for artists. But I don’t think he would be a Luddite. I can’t imagine him shunning the technology and the creative possibilities it offers.
I think he’d be in the studio playing with it, trying it out, testing it, trying to break it – looking for its limits, for the edges, the new worlds it would help him discover and explore.
I seriously doubt he’d worry about it replacing him or making his decades of work on his art redundant. I’m sure he’d see it as another tool, another technological toy that allowed him to extend his work in new directions.
As always, the technology would be a medium. The vision and the finished work would be unmistakably his.
In other words, I think he’d be a lot less hesitant than me. And assuming he knew a thing or two about creativity, then maybe I should follow his (imagined) lead, open my mind and spend more time playing around with it…
Maybe if we all had a little more faith in our own creative vision, then we would spend a little less time worrying about what this technology can take from us, and more time experimenting to see what we can create with it.
Of course a superstar like Bowie would be harder to replace than most of us. And he’d have the financial reserves to ride out any disruption. So whatever he would or wouldn’t have done, his approach may not be entirely applicable to our professional situation.
But looking at things purely in terms of creativity, I don’t think all is lost. I’m feeling a flicker of enthusiasm amid the dark clouds of uncertainty.
I’m still doing all my writing (including this article) by hand. I’m a writer. I love to write. So why would I give a robot the pleasure of doing the best bit?
But I have been using AI daily for a while now, in ways that augment my creative process – research, brainstorming, sketching and prototyping, analysis, feedback and new perspectives on my work – to see what I can learn from it, how it can deepen my knowledge, extend my skills and help me do my best work.
Dare I say it, I’m enjoying the process. New perspectives are opening up. I can see what all the fuss was about. And the more I learn about what AI can do and what it can’t, the less worried I am about AI replacing human artists and creatives.
I still have my reservations. But it turns out I’m not alone in being in two minds about AI. In his insightful book How to Think About AI, Richard Susskind writes:
‘It’s possible to be simultaneously optimistic and pessimistic about AI. It’s coherent to regard AI as a double-edged sword. This contrasts, however, with the position of many newcomers to AI – they seem to be allied to one of two camps, one that is indefatigably positive, the other relentlessly negative.’
What was it Fitzgerald said?
‘The test of a first-rate intelligence is the ability to hold two opposing ideas in mind at the same time and still retain the ability to function. One should, for example, be able to see that things are hopeless yet be determined to make them otherwise.’
So it feels like a time to acknowledge the downside, learn as much as possible about the new world, and lean into our determination – and creativity – to make the most of it.
