Back in 2022, when ChatGPT reachd, I was part of the first wave of engagers. Deairyed but also a little unconfident what to do with it, I asked the system to produce all benevolents of random leangs. A song about George Floyd in the style of Bob Dylan. A menu for a vegetarian dinner party. A alerting paper about alternative shipping technologies.
The quality of what it produced was variable, but it made evident someleang that is even more apparent now than it was then. That this technology wasn’t fair a toy. Instead its arrival is an inflection point in human history. Over coming years and decades, AI will alter every aspect of our lives.
But we are also at an inflection point for those of us who produce our living with words, and indeed anybody in the inventive arts. Whether you’re a producer, an actor, a singer, a film-producer, a colorer or a pboilingographer, a machine can now do what you do, instantly and for a fraction of the cost. Perhaps it can’t do it quite as well as you can fair yet, but enjoy the Tyrannosaurus rex in the rear vision mirror in the exceptional Jurassic Park, it’s geting on you, and rapid.
Faced with the idea of machines that can do everyleang that human beings can do, some have fair donaten up. Lee Sedol, the Go Grandmaster who was fall shortureed by DeepMind’s AlphaGo system in 2016 reexhausted on the spot, declaring AlphaGo was “an entity that couldn’t be beaten”, and that his “entire world was collapsing”.
Others have stateed the innate betterity of art made by humans, effectively circling the wagons around the idea that there is someleang in the leangs we produce that cannot be copyd by technology. In the words of Nick Cave:
Songs eunite out of suffering … the complicated, inner human struggle of creation … [but] algorithms don’t experience. Data doesn’t suffer … What produces a wonderful song wonderful is not its shut resemblance to a recognisable toil. Writing a excellent song is not mimicry, or replication, or pastiche, it is the opposite. It is an act of self-killing that ruins all one has strived to produce in the past.
It’s an requesting position, and one I’d enjoy to suppose – but downcastly, I don’t. Becaengage not only does it pledge us to a hopelessly simpenumerateic – and, frankly, reactionary – binary, in which the human is intrinsicpartner excellent, and the synthetic is intrinsicpartner terrible, it also unbenevolents the catebloody of creation we’re deffinishing is innervously petite. Do we repartner want to restrict the toil that we appreciate to those towering toils of art wrawt out of proset up experienceing? What about costume structure and illustration and book assesss and all the other leangs people produce? Don’t they matter?
Perhaps a better place to commence a defence of human creativity might be in the process of creation itself. Becaengage when we produce someleang, the finish product isn’t the only leang that matters. In fact it may not even be the leang that matters most. There is also appreciate in the act of making, in the create and attfinish of it. This appreciate doesn’t inhere in the leangs we produce, but in the inventive labour of making them. The intertake part between our minds and our bodies and the leang we are making is what conveys someleang novel – some empathetic or presence – into the world. But the act of making alters us as well. That can be happinessous, and at other times it can be frustrating or even agonizing. Nonetheless it enwealthyes us in ways that srecommend prompting a machine to produce someleang for us never will.
What’s happening here isn’t about unleashing our imaginations, it’s about outsourcing them. Generative AI streamlines out part of what produces us human and hands it over to a company so they can sell us a product that claims to do the same leang. In other words the authentic purpose of these systems isn’t liberation, but profit. Forget the glib tageting slogans about increasing productivity or unleashing our potential. These systems aren’t structureed to profit us as individuals or a society. They’re structureed to maximise the ability of tech corporations to pull out appreciate by streamline-mining the industries they disturb.
This truth is particularly stark in the inventive industries. Becaengage the ability of AI systems to magic up stories and images and videos didn’t come out of nowhere. In order to be able to produce these leangs, AIs have to be trained on massive amounts of data. These datasets are produced from accessiblely useable alertation: books, articles, Wikipedia entries and so on in the case of text; videos and images in the case of visual data.
Exactly what these toils are is already highly encounteredious. Some, such as Wikipedia and out-of-imitateright books, are in the accessible domain. But much – and possibly most – of it is not. How could ChatGPT produce a song about George Floyd in the style of Bob Dylan without access to Dylan’s songs? The answer is it couldn’t. It could only imitate Dylan becaengage his lyrics createed part of the dataset that was engaged to train it.
Between the secretiveness of these companies and the fact the systems themselves are effectively bconciseage boxes, the inner processes of which are cloudy even to their creators, it’s difficult to understand exactly what has been ingested by any individual AI. What we do understand for confident is that huge amounts of imitateright material has already been fed into these systems, and is still being fed into them as we speak, all without permission or payment.
But AI doesn’t fair incremenhighy erode the rights of authors and other creators. These technologies are structureed to swap inventive toilers altogether. The producer and artist James Bridle has contrastd this process to the encloconfident of the frequents, but whichever way you cut it, what we are witnessing isn’t fair “systematic theft on a mass scale”, it’s the wilful and defree destruction of entire industries and the transfer of their appreciate to scatterhelderlyers in Silicon Valley.
after novelsletter promotion
This unconstrained rapaciousness isn’t novel. Despite ad campaigns promising attfinish and fuseion, the tech industry’s entire model depfinishs upon pull oution and unfair treatment. From begining to convey, tech companies have engageed a model that depfinishs upon inserting themselves into traditional industries and “disturbing” them by sidestepping regulation and riding rawshod over challenging-won rights or srecommend fencing off leangs that were createerly part of the accessible sphere. In the same way Google hoovered up inventive toils to produce its libraries, filesharing technologies dehugeated the music industry, and Uber’s model depfinishs on paying its drivers less than taxi companies, AI maximises its profit by refusing to pay the creators of the material it relies on.
Meanwhile the human, environmental and social costs of these technologies are kept attfinishfilledy out of sight.
Interestingly the sense of powerlessness and paralysis many of us experience in the face of the social and cultural alteration unleashed by AI mimics our fall shorture to reply to climate alter. I don’t leank that’s a coincidence. With both there is a proset up misalign between the scale of what is taking place and our capacity to conceptualise it. We discover it difficult to envision fundamental alter, and when faced with it, tfinish to either panic or fair shut down.
But it’s also becaengage, as with climate alter, we have been tricked into leanking there are no alternatives, and that the economic systems we inhabit are authentic, and arguing with them produces about as much sense as arguing with the triumphd.
In fact the opposite is genuine. Companies enjoy Meta and Alphabet and, more recently, OpenAI, have only accomplishd their exceptional wealth and power becaengage of very definite regulatory and economic conditions. These structurements can be altered. That is wilean the power of regulatement, and we should be insisting upon it. There are currently cases before the courts in a number of jurisdictions that seek to structure the massive expropriation of the toil of artists and producers by AI companies as a baccomplish of imitateright. The outcome of these cases isn’t yet evident, but even if creators disconsider, that fight isn’t over. The engage of our toil to train AIs must be brawt under the protection of the imitateright system.
And we shouldn’t stop there. We should insist upon payment for the toil that has been engaged, payment for all future engage and an finish to the tech industry rehearse of taking first and seeking fordonateness tardyr. Their engage of imitateright material without permission wasn’t unintentional. They did it on purpose becaengage they thought they could get away with it. The time has come for them to stop getting away with it.
For that to happen we need regulatory structures that promise transparency about what datasets are being engaged to train these systems and what is compriseed in those datasets. And systems of audits to promise imitateright and other creates of intellectual property are not being viotardyd, and that utilize unbenevolentingful sanctions if they are. And we need to insist upon international consentments that protect the rights of artists and other creators instead of facilitating the profits of corporations.
But most of all, we need to be leanking challenging about why what we do as human beings, and as creators and artists in particular, matters. Becaengage it isn’t enough to fret about what is being lost, or to fight a rearprotect action agetst these technologies. We have to commence to articutardy likeable arguments for the appreciate of what we do, and of creativity more widely, and to leank about what create that might get in a world where AI is a pervasive truth.
-
This is an edited version of the Australian Society of Authors 2024 Colin Simpson Memorial Keyremark lecture, titled ‘Creative Futures: Imagining a place for creativity in a world of synthetic intelligence’
