Zack London, a 33-year-old American artist, is renowned for his distinctive online persona – Gossip Goblin, followed by nearly half a million people. His Instagram followers recognise him from his quirky adventures of his invented character, Fredvog. This effigy, a virtual explorer dressed reminiscently of Gandalf adorned with a red gnome cap, engages viewers in a world of peculiar, Tolkienian beings: from hobgoblins, mini communities inhabiting wild fungi to supernatural creatures from the netherworld, a sacrosanct cat kingdom and their long-time rivals, the mole people.
The narratives London concocts may seem dreamlike yet are weirdly fascinating. What sets his work apart is the complete reliance on artificial intelligence (AI) to put together everything from his calming voiceovers to his realistic visuals.
In his Stockholm-based residence, he explains that this is a complete departure from his previous work. He possesses an astute understanding of tech progress as a User Experience (UX) designer. A few years ago, he discovered that a machine could output comparable quality artwork as his, which would usually take over a hundred hours. This realization inspired him to play around with generative-AI applications.
These are artificial intelligence types that design content such as text, music or graphics by studying patterns from prevalent datasets, then concoct similar material based on user prompts. It’s worth knowing that ChatGPT, the world’s most utilised generative AI, gets over 1.8 billion website visits a month, more than Netflix.
To generate and bring his video graphics to life, London employs a combination of economical Midjourney and Runway software, which he matches with his scripted narrative rendered in a machine-made voice. While his online content is not monetised, he has sold commissioned work via his website. Over the last year, equally whimsical projects mashing popular culture entities with eccentric twists have brimmed the internet. Viral videos featuring Harry Potter characters in a Balenciaga-themed sequence have garnered 12 million views, rallying 4 million views for Wes Anderson’s take on Lord of the Rings. This rudimentary yet playful creation poses a significant challenge to the conventional art and entertainment sphere.
London, a renowned AI artist, reveals that the application of AI reduces those the complex processes involved in creating art, altering the perception of craftsmanship in the industry. It highlights that the notion of expertise in ‘prompt engineering’, as claimed by many AI artists, is false, as the generation of AI images does not necessitate any real ability.
Meanwhile, Yining Shi, the principal research scientist and senior engineering manager at Runway, suggests that predicting AI’s future trajectory remains challenging due to rapid technological advancements. Yet, these developments could simplify professional media production, granting individuals the liberty to produce intricate content effortlessly and flexibly. This ‘democratisation’ in media production could subsequently trigger a spike in creative output and personal expression. Runway’s ambition is to allocate more time and resources to film-makers and artists by emphasising ‘human augmentation’ rather than complete substitution, a concern prevalent within the industry.
This erosion of technical barriers excites many, but also sparks anxiety among industry professionals, particularly animators, whose skills appear to be overtaken by AI’s generative ability.
Jeffrey Katzenberg, co-founder of DreamWorks Animation, the powerhouse behind legendary films like Shrek, Kung Fu Panda and Madagascar, warned last year that AI might soon replace up to 90% of animation artists. He nostalgically referred to the times when producing a top-notch animated film required 500 artists and five years, and predicted that in the near future, this could be reduced to barely 10% of the current requirement. However, he remained optimistic that those who survive the technological transition will require creative abilities to navigate the software, turning prompting into an artistic asset.
The ripple effects of AI’s integration are palpable already. OpenAI, known for developing ChatGPT, recently unveiled Sora, its new video creation instrument. Using straightforward text prompts, Sora can produce life-like videos that can deceive even the most discerning viewer. The tool hasn’t been made publicly accessible yet, but its revelation has generated mixed responses of anticipation and apprehension.
Barry O’Sullivan is an AI and ethics expert who serves as a professor at University College Cork. He believes that the perceived negative impacts on the art and entertainment sectors have been greatly exaggerated. While he admits that AI will lead to changes in certain jobs, some quite drastic, he does not see widespread unemployment as a result of these alterations. Drawing on commentary from Katzenberg, he puts forth the idea that the demand for animation might actually distribute work around numerous small-scale projects instead of the once popular DreamWorks blockbusters.
Contrary to the predicted mass layoffs, Professor Sullivan explains that they have not materialised as expected. Instead, during this period of technological advancements, we’ve seen the sprouting of exciting new positions, novel industries, and fresh professions.
As far as the quality of AI-created art is concerned, there’s no denying that AI can produce content at an impressive speed, a task that would typically require months for artists. However, there’s an ongoing debate about the merit of such works and whether they can hold their place alongside traditional art. This argument flared up when David Lester Mooney, an artist, used AI to design an image of four young women dressed in 19th-century clothes, named Throwback Selfie #Magdalene. The artwork was chosen for the Royal Hibernian Academy’s Annual Exhibition, triggering an online storm, with some dubbing it “grotesque” rather than art.
In a similar story, a work called Théâtre d’Opéra Spatial, which presented a space-opera scene reminiscent of the renaissance, nabbed the grand prize in the digital-art segment at the annual art contest of Colorado State Fair. Some critics declared it signaled the end of classic artistry in the AI age. A couple of years down the road, though, such judgements seem excessively dramatic based on one minor category of a relatively minor competition.
Zack London isn’t quick to applaud those who generate creative content using AI technology. He believes that the ease with which breathtaking imagery can be produced doesn’t warrant recognition. A case in point is the proficiency of Midjourney, where a simple prompt such as ‘girl’ can generate an artwork akin to a renaissance masterpiece. However, London feels people should not be praised for such outputs as if they were a testament to their own artistic abilities.
London resents the concept of individuals branding themselves as AI artists. He argues that simply inputting search terms, like ‘hot girl’, into Google does not qualify someone as a software developer. He suggests that as creative production becomes more streamlined, the focus should pivot towards storytelling and originality. Decrying the label of an ‘AI artist’, London insists he’s proven his chops in the traditional arts, prior to the advent of AI.
With mixed feelings regarding AI’s potential in the realm of art, London contemplates artistic value by drawing comparisons between a Rembrandt classic and Levitated Mass, a 2012 installation at Los Angeles County Museum of Art featuring a 340-tonne stone, the installation of which, came with a staggering price tag of $10 million. Despite the contrast, he concludes that the evaluative process is not a matter of comparing apples to oranges, rather, AI sits in its own anomalous niche within the broader artistic context. It’s a nebulous space yet to be defined appropriately, a zone that might not even fit under the conventional term ‘art’.
As a member of the Government’s AI advisory council, Prof O’Sullivan holds a confident stance, stating that generative AI is not seen as artistic in the AI community. He points out that AI systems lack worldly understanding, so although they can generate content, there is no real cognition behind it; essentially, they’re just connecting words and phrases with associated data. In his view, true art reflects the artist’s insight, interpretation, and commentary on their surroundings. However, he also predicts a rise in artists utilising AI for crafting bold statements about the technology itself.
Mary Cremin, the programming lead at the Irish Museum of Modern Art, notes that the embrace of AI isn’t uncommon in the art community. Artists have incorporated technology into their work, evidenced by Jon Rafman’s algorithmic creations and Doug Aitken’s innovative 360-degree video piece that includes an AI-constructed choir. Furthermore, renowned Irish artist John Gerrard utilises digital simulations. She observes a surge in interest for digital art, especially amongst the digitally literate generation, though it remains inherently subjective when determining what constitutes as ‘good’ art.
Then there’s the matter of copyright, an issue yet to be comprehensively addressed. AI, largely unregulated globally, trawls digital datasets to gather images and text for its global understanding, often using copyrighted material for generating content. This gives rise to enquiries concerning the ownership of the final product. The European Union has attempted to regulate the development of AI through its AI Act. As Prof O’Sullivan puts it succinctly, while artists may draw inspiration from various sources, when does that inspiration cross the threshold into intellectual property infringement?
“A challenging question concerning copyright infringement and AI is brought up by Barry Scannell, an AI-law expert at William Fry and also a member of the Government’s AI advisory council. Scannell states that there could be incidents where copyrighted works, utilised in training AI systems, might be deemed as breaking copyright laws. He also mentions ongoing international lawsuits alleging copyright violations.
The uncertainty extends to the content produced by AI. According to Scannell, UK and Irish copyright legislation suggest that the instigator of the generation of a computer-created work, in absence of a human author, can be regarded as the creator. However, comparing this standpoint with the cumulative copyright laws and regulations of the European Union, Scannell expresses doubts over its potential to withstand an appeal to the Court of Justice of the European Union, believing it deviates substantially from EU copyright laws.
Scannell underscores that this subject is still very much evolving in jurisprudence, raising unexplored questions, and definitive answers remain elusive at present.
The rapid progress of AI may give it an uncivilised vibe to AI engineers. AI’s inevitable encroachment, whilst a cause for optimism among artists such as Zack London, foretells major upheaval in the arts and entertainment sphere. From his perspective, technological innovation doesn’t always obsolete past practices. Photography didn’t extinguish the art of painting but altered the perception of has its own unique value and cannot be replicated by technology,” London observes.
This situation brings to mind the historical events of the Luddite uprising. Their actions, which included setting fire to looms and early manufacturing establishments hence resulting in the weavers’ unemployment, seemed logical at the time but, upon reflection, appears rather ridiculous… Displaying a tag of ‘No AI’ on your online profile, it comes across as a feeble attempt similar to a lone ant taking on a mountainous rock. Indeed, it’s a messed up situation. However, the world we live in is chaotic and this occurrence aligns with the unavoidable course the world is taking. Governments will never fully control this. It is something that exists in the world we live in. No level of emotional appeal will hinder its progression.