Calculated Creativity
As AI gears up to create, how can culture preserve its ability to enrich?
Tasked with future-proofing the company’s global video businesses, Telefonica Group’s Media & Entertainment Innovation team is exploring how technologies like VR, AR and AI are changing how stories are created and consumed — and how these changes could, in turn, affect audiences and fans. This post focusses on what we have learned about AI in the context of culture creation so far.
“If it got to a point where a robot was screenwriting and it wasn’t good, then that might turn people off storytelling. The entertainment industry needs good, original stuff — otherwise it’s going to die and go away.”
— Max (Film Creator/World Builder/VR Enthusiast, Los Angeles)
A New, Old Idea
Like many ideas concerning the latest innovations, ‘Calculated Creativity’ is not a new one. Ulterior motives to artistic ones, such as power signalling or commercial gain, have dominated culture creation ever since ancient Mesopotamian kings commissioned sculptures of their likeness to cement their authority, inspire admiration, and foster subservience.
The histories of the visual arts, music, literature, and entertainment all offer up endless case studies of ‘Calculated Creativity’. Renaissance artists predominantly served to enhance the prestige and status of powerful commercial and religious leaders, over and above artistic expression. In music, we thank patronage, not album sales, copyright laws or sold-out world tours, for Classical’s ‘greatest hits’. The genesis of television entertainment in the fifties, the soap opera, came into existence as a means to sell its namesake.
We now find ourselves in a moment in time where a new participant — artificial intelligence — has joined the age-old dance between creativity and commerce across the cultural spectrum. Some consider it a powerful new creative tool, while others lament it will irreparably disrupt and destroy culture. Either way, AI will continue manifesting as a creator of culture and storyteller in multiple ways — in film and television, in art, in literature, and music.
Optimising Culture
AI, at its core, is about optimisation and efficiency. In industries like manufacturing, finance or IT, it is applied to do more, better, faster — and, crucially, for less — than any human doing the same job. Cultural industries are now following suit, exploring applications of AI that can enhance creation and consumption processes.
But what happens when you apply a technology that seeks to optimise — quickly and cheaply — to cultural industries? Doesn’t the sheer premise of AI’s capabilities jar with the output of industries that ‘manufacture’ enriching, moving, thought-provoking ideas, conceived by individuals highly skilled in turning their imagination into experiences? What will it create, and how? Will it be any good? Under what conditions and categories can AI-generated culture legitimately exist? And, crucially — what will those at the receiving end (i.e. the audience, the viewer, the reader or the fan) make of it? Will they even be able to tell or at least feel the difference? Will it matter?
Before we can attempt to answer these questions, let alone map out implications for innovation, let us first look at how exactly AI will manifest in the creation and consumption of culture (specifically, though not exclusively, in entertainment).
Manifestations of AI in Culture Creation
1. Discovery Engine
Platforms dedicated to cultural content (mainly music and film/television) are continually seeking to improve their algorithmically-powered discovery engines. Along the way, they fragment audiences into nano-clusters, ripe for being targeted with content in line with very specific tastes, modes and moods. These increasingly sophisticated engines don’t just enable audiences to find cultural content — they have also encouraged millions of creators to emerge from small/specialist/obscure subcultures, allowing them to find, nurture and monetise (suddenly) substantial audiences for their output too.
2. Creator’s Mindset
This manifestation relates specifically to image, music, and video creation on creative social platforms (Youtube, Instagram, Tiktok), whose respective algorithms are known to reward streaks of sustained, voluminous hits. This has resulted in said algorithms arguably hijacking creators’ collective mindsets, prompting them to optimise their output for high returns on engagement over creative expression.
“We know the audience. We know what they’re expecting…That’s why our shows are about relationships with attractive women. And showing abs.”
— Karina (Amateur YouTuber, Los Angeles)
Mainstream artists, screenwriters for networks, Hollywood studio executives and pop music producers have arguably operated under similar conditions for decades — albeit without as many real time metrics or the motivation of vast ad-dollar powered riches that digitally native creators have access to today. One notable consequence of the algorithmically-hijacked mindset is the deployment of extremes to game algorithms for maximum exposure and financial gain.
Creative expressions or stories that entail danger, psychotic outbursts, outrage, extreme beauty ideals, hyper-sexuality or flaunting extreme wealth are lucrative currency, given their ability to trigger heightened emotional responses and resulting in more engagement. The consequences of repeat exposure on the audiences’ collective psyche entails being given permission to normalise, aspire to and/or emulate it.
“You can have certain experiences in this world, they produce certain desires, those desires reproduce our world.”
— Moxie Marlinspike (Signal Founder, as cited in The New Yorker, 2020)
3. Optimisation Tool
In film, music and online literature, data has established itself as a crucial ingredient in the creation process. In order to democratise the capabilities Netflix and Hulu pioneered, a whole new industry around data-driven production is emerging to facilitate this approach to culture creation. Its pitch to producers, creators and distributors alike is to choose data over gut instinct in order to maximise their return on investment. In the entertainment field, companies like Script Book and Cinelytics are leading the way.
Positioning themselves as an interface between creativity and technology, they provide real-time analytics around budgets, genres, ratings, box office performances of directors and cast members, talents’ social clout (which plays an important role in marketing new ventures), even story themes and moods. On a dashboard, multiple ingredients for stories can be swapped in and out, or dialled up or down, until a satisfactory prediction of profit is eventually generated. Debates are rife concerning the quality of these so-called ‘data-pull’ shows, with some industry insiders deeming them to be ‘soulless’ and ‘neutered’.
“It feels like we’re entering into an age of massive mediocrity…there’s an unbelievable amount of ‘just OK’ out there.”
— Financial Times (Whatever Happened To The Golden Age of TV, March 2020)
“It’s all remakes and franchises now…one week at the cinema, the marquee was all reboots. When Jordan Peel’s Get Out came out, that made me happy — an original thing came out, and it made money. If AI hurt that, people wouldn’t go back to it.”
— Max (Film Creator/World Builder/VR Enthusiast, Los Angeles)
4. Creative Catalyst
Artists and creators experimenting with AI have unanimously described the technology as a creative catalyst that pushes them beyond their familiar comfort zones, thereby evolving their canons in new ways. Trained on words, tunes, movements or images created or selected by artists/creators, AI (in the form of generative software) can generate new expressions of their creativity — all the while the artist remains in control and decides what does justice to what they want to express. With brain-machine interfaces and wearable MRI’s that tag-team with GAN’s (generative adversarial networks) on the horizon, culture creators may one day be able to ‘think it to film it’, without having to apply craft, time or tools to create artefacts of expression.
5. Responsive Storytelling
“Personalising a film to someone’s tolerance could be interesting — so, in horror, one person gets more gore than another, based on what they can handle.”
— Karina (Amateur YouTuber, Los Angeles)
As technologies that track instinct, personality, and emotion continue to evolve, cultural output will increasingly become tailored to the audience’s needs or preferences in real-time too. Virtual storytellers and entertainers (either synthetically created, or avatars of real-life performers) will become highly attuned to their audience-of-one — specifically, to the meaning of every reaction, thereby enabling them to deliver new heights of engagement. Motion/emotion sensors and computer vision can facilitate compelling environmental experiences too, tempting physical expression and triggering engaging responses.
It is, undoubtedly, compelling for creators to understand how to push (and monetise) their audiences’ buttons via rich troves of data in this context. As the founder of a responsive story company revealed: “We’re hiring more and more data scientists to work alongside writers. We can see that people really engage with highly intense moments that trigger strong emotions, like shock. So we can evolve the story to give them more of what they want…what keeps them engaged.”
Paired with an alleged onslaught of virtual or synthetic beings (one experts predicts we will be surrounded in the next 10–15 years), regulation or ethics codes of practice are needed to ensure creators’ intentions are benign rather than exploitative. In order for them to play a sustainable part in storytelling’s future, responsive stories/beings must be aligned with positive intents and purposes, to mitigate isolation and division — not least because the notion of a shared reality is already diminishing in the current digital media ecology.
In light of culture’s vast potential to connect us around shared passions, it remains to be seen how much of our bandwidth such ‘audience-of-one’ experiences will claim in the future. One thing is certain — it will bring audiences and fans closer to the stories, worlds and characters they (sometimes lightly, sometimes deeply) care about. Responsive storytellers across all cultural forms must, however, also be mindful of the wider context of consumption. What might the socio-cultural repercussions look like when sharing mutual experiences of culture are no longer currency, and opportunities to connect with others over passions and interest disappear from social interactions?
6. Autonomous Creator
Various experiments are underway across many forms of culture, training the first generation of autonomous AI culture creators. There is a slight fallacy in the notion of autonomy — robot painters, software-based fiction writers, and synthetic composers will still need to be trained on existing content data generated by human creators data to kick-start their creation process. As we wait to consume more than just the experiments coming out of various labs, we can at least start asking ourselves what role(s) autonomously-created stories can assume in culture — and what we want it to assume.
Will AI produce the chicken nugget equivalent of culture — a low/no-nutrient filler that briefly hits the spot (yet fails to truly satisfy, let alone be memorable)? This would be the purest translation of AI’s abilities in the culture-creation space — churning out vast quantities of stories, images or tunes cheaply, for maximum returns, akin to the fast food model. Yet just like the latter, this may potentially result in a(n even greater) division between those who can afford more ‘nutritious’ culture, and those who don’t have any other choice but to consume processed ‘junk’.
Or will AI produce cultural output that functions purely as a mood shifts we call on and consume whenever we want to feel differently? What about producing average, cheap forgeries that appear to be good enough and therefore better value — the cultural equivalent of a counterfeit Chanel bag perhaps? What are the implications for the intellectual property of the creators who generated the originals in this scenario?
Might AI-generated culture prove the sceptics wrong, and manifest as a critically acclaimed “new sub-genre of culture” (as pondered by artist Hito Steyrl, in conversation with Hans Ulrich Obrist in Possible Minds, 2019)? The answers are, probably, yes all round — yet it remains to be seen which ones sustain the transition from tech-driven experiments to in-demand cultural experiences/stories people will want to engage with, crave more of, and pay for over time.
Tech Possibility Vs Human Truths
“Music is feeling. It’s impossible for a machine to feel things. Music is more than a melody. I wouldn’t follow an algorithm artist. But knowing a real person wrote the music, that makes a different. Then the technology is adding, not replacing.”
— Daniel (Music Fan, Madrid)
There is, however, one particular human truth that can’t be ignored by those who are intent on pushing the boundaries of tech-led creation. More often than not, familiarity with a culture creator’s life, their values, creative influences and viewpoints enriches the experience of their creations, across the cultural spectrum. This poses the question of whether culture can still connect with us on a deep and satisfying level when you take the creator’s humanity (including their uniquely processed experiences) out of it.
The film-maker Alfonso Cuarón drew on his experience of having a strong maternal bond with his family’s housekeeper to write Roma, and in turn created something that deeply resonated (against a lot of odds/despite subtitles) on a mainstream streaming platform. The musician Nick Cave’s work connects in moments of despair because his music was inspired by personal loss his listener identifies with. The pianist and composer Chopin was perpetually lovesick (as well as physically sick with syphilis), which robbed him of his physical and emotional strength — thus resulting in some of the most delicate tunes produced for piano. For committed audiences and fans, this kind of context elevates the experience of culture from one that is transactional or functional, to one which is rewarding and memorable.
Perhaps it is Parasite’s recent win at the Oscars (and its impact on audiences world wide) that is testament to the ongoing importance of artistry, imagination and creativity in a world looking to technology to make more from less. For the sake of not losing the deeply satisfying enrichment audiences/fans draw from culture across the high-low spectrum, these qualities must be actively protected and championed.
Dance With Caution
It is clear that AI is quickly establishing itself as a new player in culture creation and storytelling. It has the potential to accelerate creation and distribution, but also to override human imagination. It could render human inputs into art, film, literature, or music redundant, in favour of higher volume and predictable, lucractive returns. On the flip-side, it may also conjure up new genres of culture and storytelling when utilised as a creation tool (responsibly), and fundamentally change how we engage with culture and stories in the process. We can cautiously ascertain that as one end of the industry becomes fixated on high-performing quantities of culture aimed to offer a ‘fix’, the other end will push exceptional quality that can deeply enrich.
When looking to AI as a potential ingredient in their creative processes or businesses, content creators, as well as their related industries and governing bodies alike (e.g. cultural institutions, government, trusts, unions, etc) should consider three implications in their pursuits. Not only will these implications minimise the threat of letting a force driven by optimisation of returns diminish culture’s collective ability to add value to our lives, cultures and society — they will also ensure technological possibility is aligned with human truths.
1) Intention
This predominantly concerns the creator’s relationship with their audience. Creation should ideally start with a desire to tell a compelling story, or expressing a compelling idea, grounded in an understanding of the prospective audience. Who is this for? Why will they care? It is important to consider what constitutes a connection with the audience at the heart of the creative process — so avoid using the technology for technology’s sake, let alone creating culture purely as a way to grow the bottom line.
2) Collaborative Experimentation
Experimentation that fuses technology and culture creation is best carried out by multidisciplinary teams. The ideal blueprint entails entrepreneurial spirit, engineering prowess, data science skills, creativity and imagination, and expertise in the humanities to answer key questions about role and meaning.
3) Regulation
Akin to France’s ‘Cultural Exception’ policies (e.g. 40% of music on the radio to be by French artists, 20% of which must be new), cultural departments and content industry governing bodies should consider setting out parameters for the quantity of production utilising AI in entertainment and factual content development. Furthermore, standardised disclaimers will ensure the audience is made aware of why they’re seeing what they’re seeing, thus enabling them to decide if they want to engage or not.
“What we are actually listening to is human limitation and the audacity to transcend it. AI, for all its unlimited potential, simply doesn’t have this capacity…If we have limitless potential then what is there to transcend? And therefore — what is the purpose of the imagination at all…Where is the transcendent splendour in unlimited potential? […] AI would have the capacity to write a good song, but not a great one. It lacks the nerve.”
— Nick Cave (Musician, Red Right Hand Files, 2019)
