Art Nemesis
In recent weeks, my social media feed has been prompting me with several ads for AI powered content generation tools. I don't remember ever googling that rather specific thing, so perhaps the first ad appeared at random... though “random” is not really an appropriate term these days. Everything we see is data driven, so whilst I might not have gone searching for “AI powered content generation tools”, my online activities have algorithmically still delivered that result to my door. It occurs to me in this moment that researching a recent article about procrastination and its place in the process of writing might have suggested to the machines that I'm a potentially easy mark for this service. The reason I noticed that first ad (if indeed it was the first), amongst the many suggested posts I subconsciously discount was the accompanying picture, incongruous as it was with the text, but of a face familiar to me – a person whose YouTube channel I have occasionally visited. Cynically discrediting the idea of AI generated prose (Surely it could never swear as creatively as me!?) whilst rebuffing the clickbait tactics of the “randomly” assigned image (presumably a machine-chosen lure curated for different users...) I moved on. Although I never actually clicked through to anything on offer, the fact I slowed my scrolling enough to not just acknowledge the ad, but implicitly tell the algorithms I might want to see more of the sort seemingly validated the machine-learning on display.... cue many more ads of the sort.
Ironically enough, AI generated content has, by way of a targetted, algorithmically-chosen ad, generated content – this content... just not revenue for the advertisers. Nevertheless, it's testament to how sophisticated things are becoming. A jaunt down my feed this morning brought a related ad in the form of MovieBot, an AI powered tool which promises to turn your short TV or movie ideas into fully fledged scripts. Having once read the hilarious (yet unnerving) rally address generated by an AI bot fed 1,000 hours of Donald Trump speeches (Google it), my expectations for coherence were not high. What passed freakishly plausibly for Trump-speak would surely not work for anyone with a normally functioning brain? Nevertheless, amongst the many dismissive comments deriding both the presumed efficacy of the tech and the current state of cinematic storytelling as soulless even before removal of the human element, were several surprisingly enthusiastic responses. Some of these were from actual users and others appeared to be from entertainment industry professionals (or very convincing moles), disagreeing with the cynics and heralding the tech as another sandbox to play in – a new tool to strengthen rather than supplant a scriptwriter's arsenal. Far from signalling the end of human creativity, could it not represent a new avenue for blue-sky thinking? Perhaps writer's bot might help alleviate writer's block?
An adjacent debate seems to be in a state of flux concerning much of AI implementation, between those who see the benefits and those who fear the repercussions. A field which has always stirred debate is art (and what is art?) and here also, inroads are being made by AI, stirring the pot to predictably polarizing effect. 2022 saw the explosion of several programs such as DALL-E 2, Stable Diffusion and MidJourney which use machine learning and the scraping of vast data sets to create novel images based on text prompts, image references and parameters suggested by users. The relative quality of these images varies dramatically, but when the process works, it's astonishing. Certain users have gained swift recognition for their ability to effectively prompt the tools into creating evocative, realistic, consistent or cohesive works. Galleries have already come knocking.
Critics still admonish the apparent lack of creative input and numerous red flags have been raised in terms of copyright issues and the effect of the technology on a community of “traditional” artists already struggling to make a living in a world which never fully appreciates how much more vibrant and interesting a place it is thanks to their input. Many are now decrying AI art as not art whilst simultaneously declaring “Art is dead”, a hysteria frighteningly similar to when the proliferation of photography in the early-mid 19th Century had painters and critics up in arms. Nevertheless, the tech's methodology is concerning, as the programs scour millions of images from the open web, teaching algorithms to recognise relational patterns to the end of recreating certain aesthetics and styles in newly generated work. By uploading their art to the net, human artists are arguably training the instruments of their destruction to better imitate and potentially supplant their intellectual property.
Not everyone agrees we have a crisis, provided companies commissioning AI art and individuals creating it stick to a necessary code of ethics.... Good luck with that..
Some celebrate AI art as the most prescient form of art there is – ultimately it's the art of big data and we live in an age of big data. Its ability to quickly straddle, augment and reconfigure vast information sets far outstrips our own.
Many argue that everything's essentially already been done and there's no shame in simply modifying. Louis Vuitton's late artistic director and multimedia powerhouse Virgil Abloh was quoted as saying “A creative only has to add a three per cent tweak to a pre-existing concept in order to generate a cultural contribution deemed innovative.”
When critics accused him of artistic plaigarism, his retort: “Duchamp is my lawyer.”
Marcel Duchamp's 1917 piece “Fountain”, a signed urinal, was arguably the first piece of conceptual art and we've been arguing about what constitutes art ever since, it's just now we can argue with machines.