Is AI killing creativity?

OK, cards on the table, I’m worried about AI. Really worried.

As a writer and editor I’m worried that it’s steering my career path towards obsolescence. As a human I’m worried about its environmental impact and its lack of ethics and regulation. And as someone who works in the arts, I’m worried about what it means for creativity – both for the artists and the art that is produced.

But am I right to be so concerned? Is AI actually killing creativity or am I just being close-minded? Because AI is a broad term, and to solely align it with slop social media posts, terrible takeaway flyers and deep fake videos of Andy Burnham and Sadiq Khan doing the hokey cokey (OK, bad example, I would watch this) is to ignore its true potential. AI has already been of significant benefit to people with access needs, and to the wider health service, and if it can free up doctors, nurses and care-givers to be their compassionate best, then surely it can also free up artists to explore new creative avenues.

The impact and potential of AI on creativity is just one of the topics set to be explored here at the Southbank Centre this September as part of Creative Intelligence, a weekend of events and talks exploring the new ideas, art and technologies that are shaping our future. As someone who works at the Southbank Centre, the imminence of Creative Intelligence, also offered me an opportunity to try to allay my worries about AI, by getting in touch with people much better versed in its potential than I am. And reader, I took it.

Well, not straight away. Initially I typed ‘will AI kill creativity?’ into Google. The search engine’s AI overview instantly assured me that ‘AI will not kill creativity’. But then it would say that wouldn’t it. Better then to take advantage of available human insight, and so I reached out to three creatives who’ll be joining us at Creative Intelligence; Bea Taylor Searle, a cultural producer and co-founder of Peckham Digital, which is dedicated to widening access to creative technology in the arts; Justin Spooner, co-founder and director of Unthinkable, a digital strategy agency with a focus on how digital technology can better serve people; and Razik Darji, a creative technologist who blends emerging tech, culture, and storytelling to build award-winning digital experiences.

AI is no longer something of the future, unlike hoverboards and roast dinners in pill form it’s already here, so are this trio already indulging? Darji certainly is, and is keen to stress how useful it’s proving. ‘I’m neurodivergent,’ he states, ‘sometimes I can see an idea far more clearly than I can explain it so AI helps me translate what is in my head into something other people can see, understand and believe in.’ Alternatively, Taylor Searle, tries ‘to steer clear from use of AI’ in her creative practice, while Spooner sits somewhere in between. ‘My role at Unthinkable means I need to try and understand where there are new methodologies and approaches to be discovered’, he explains. As such he’s found himself ‘doing a lot of experimentation with AI’, including using it to ‘help articulate [his] own thoughts and draw in different ideas’.

Of course using a thing doesn’t mean you’re necessarily immune to concerns you may have about it – public wifi, self-serve coffee machines in petrol stations, the Bakerloo line (why so screechy?) – so, do any of these experts share my worries? The answer it transpires is yes, very much so, and for Taylor Searle that starts with how the language we use around it skews perceptions of what it is. ‘The name itself, ‘artificial intelligence’, is problematic,’ she says. ‘It anthropomorphises the tool from the beginning, automatically creating an us-versus-them situation, which is then augmented when we talk about it thinking and in cases describe it as ‘creative’’. She suggests that referring to it instead as a ‘cybernetic tool’ ‘might support moving away from that anthropomorphic problem’.

Spooner’s worries sit more with how AI is established. ‘I’m concerned that AI systems trained on existing culture may encourage sameness while presenting it as limitless possibility,’ says Spooner. ‘I think this might be a general issue with how large language models (LLM) and the tools we use to access them have natural limits, or how they are not actually designed well for creative work.’ Spooner also admits to being ‘worried about the concentration of power behind these tools and the effects on creative labour, on specialist creative disciplines, and the questions that arise around intellectual property ownership’.

‘There is an assumption that creativity should always become faster, easier or produce at larger scales,’ continues Spooner. ‘Art also depends on difficulty, ambiguity, independence and time to explore – these are qualities that may have cultural value even when they resist classic economic measurement. Something about internet culture and its insatiable appetite means we tend to see Generative AI through that lens, one of optimisation or productivity’.

The impact AI’s advancement and wider use may have on artists is a concern shared by Darji. ‘The same technology that gives more people the power to create can also take power away from the artists whose work it was built on,’ he says, adding ‘we need much stronger conversations around consent, attribution and compensation, because democratising creativity cannot mean exploiting creatives’.

Darji also shares Spooner’s worry about the ‘concentration of power’ behind the technology. ‘We do not talk enough about who gets to shape AI. If the rooms building these systems lack diversity of culture, class, disability and lived experience, entire perspectives can be misunderstood or excluded before the technology even reaches us’.

And on top of the worry about AI misrepresenting or excluding perspectives in its creation, there is also the issue of the resultant tools and applications being used nefariously. ‘The spread of misinformation through AI technology is incredibly concerning,’ says Taylor Searle. ‘We live in a moment when the president of the USA thinks it is acceptable to litter the internet with AI-generated images of his policies. There are people who understand that type of visual representation better [than policy discussion or reporting] and so become fueled by false narratives through these generated images. This is a real threat to future democracy’.

So far, so terrifying. While on one hand there’s an element of reassurance to be gained from knowing people closer to AI have similar concerns to mine about its impact, the other hand is desperately pulling the duvet back over my head. But it can’t all be bad, can it? Most new technologies land on our table with a side order of fears, and many have ultimately proven to be unsubstantiated – television didn’t make the radio obsolete; the camera didn’t stop people painting; the internet didn’t kill off the printed word. So, what’s the good news? What can AI do for our creativity?

‘AI can democratise creativity by shrinking the gap between having an idea and being able to make it,’ says Darji. ‘For people without huge budgets, networks or resources, it can turn imagination into something tangible before anyone has to give you permission’. That’s more like it. And it doesn’t stop there. As well as levelling the playing field for artists, it can also give them greater freedom to be creative as Taylor-Searle enthuses. ‘Currently artists must navigate a landscape of applications – to funds, residencies, exhibitions – while also holding a job to support their livelihood, before they can focus on their practice.’ Tasks which she believes AI can play a big part in relieving; ‘It can be a good tool for admin tasks, for updating budgets, managing finances and re-writing applications to fit different formats, freeing up time for artists to be artists’.

For Spooner the excitement about AI’s potential for the arts sits in how it can open up new creative processes. ‘My sense is that with AI, experimentation can become more accessible,’ he says. ‘There are some aspects of how LLMs work that make me think about processes like evolutionary design practice, where we create a wide range of variant ideas which we then select from, and evolve through, a number of iterative creative processes. AI tools can help us develop and extend our thinking and also help us to move between ideas, perspectives, and forms in a more fluid way. In AI you can have a tool that acts almost like a reflective partner to help you potentially get to ideas you wouldn’t on your own’.

For examples of AI enabling experimentation to become more accessible we don’t have to look far; it’s something that Darji has already utilised in his own practice. ‘AI lets me attempt things I previously would not have had the money or technical resources to build,’ he explains. ‘I created an app that helps my Deaf mother experience music differently, and that is what excites me most: suddenly the distance between an idea and making it real is much smaller.’

This idea of leaning on AI to help bridge existing gaps also holds appeal for Taylor Searle who is ‘excited by small AI, small datasets that artists might put together and train locally’. As she explains, ‘small AI has the possibility of providing new approaches to one’s practice. For example, how would manually tagging your work for a computer to process it change the way you see it? How does prompting then become a tool in a creative practice? If there has to be a future for AI in creative practices it is built locally, with selected data that has the appropriate consent and consideration within it.’

Which opens up another facet of AI usage I hadn’t previously considered. Because it’s easy to see AI in its broadest sense as being solely owned and controlled by ‘the man’, which is why Spooner and Darji echoed my concerns over who is driving it forwards. But is that actually the case, are we just at the mercy of the whim of ‘Big Tech’ or can artists and creatives themselves influence the future of AI?

‘Artists can definitely influence the future of AI,’ Taylor Searle reassures me. ‘Artists are able to critique AI and ensure it works for us rather than consume and exploit us. This is why they should be more involved in the process of building these tools and be in the same rooms as the engineers and the programmers; artists will provide the questions that the engineers might not consider during development.’ Darji is in agreement; ‘Technology companies may build the tools, but artists shape the culture, meaning and behaviour that forms around them, so creatives need to be in the room helping decide what this technology becomes.’

Wherever this room is, it’s getting crowded, but Spooner, like Darji and Taylor Searle, is equally convinced that its artists who are most likely to ask the important questions and should have the opportunity to do so. ‘Artists, as always, can reveal what these systems overlook; challenge the values built into them and imagine relationships with technology that are idiosyncratic and don’t have to follow the prevailing paradigm. And this is partly what Creative Intelligence is about, it’s about the role of artists and creatives being not merely to find inventive uses for AI, but to help society decide what we want it to be, and to show us different ways to think’.

So, where does this leave us? Rather than being a threat to creativity, is AI actually more of an opportunity for it? ‘It’s both,’ says Spooner, quashing my attempt at easy pigeon-holing, ‘but neither outcome is inevitable. Each creative technology presents a similar dilemma. It takes many people lots of thinking and doing to find new ways to work and determine how our ethical frame needs to be retuned. When artists can interrogate, critique and reshape it, AI becomes an opportunity. But when only commercial logic determines what creativity is allowed, it becomes a threat.’

Taylor Searle is similarly pragmatic. ‘AI has become a reality we will live with until I die and beyond,’ she states, adding, ‘the thing that will decide if it is an opportunity or a threat is how we build our relationship with it. This entails having a level of agency over it, rather than strictly following the rules for its use provided by Big Tech. Opportunity lies with the artists who disrupt its use to develop healthier relationships to it.’ ‘For me, it is an opportunity,’ adds Darji, reiterating Taylor Sarle’s point. ‘I left school at 16, grew up working-class in a council flat, and AI has helped collapse barriers that once required money, credentials or entire teams to overcome.’

Whether we like it or not, whether we feel ready for it or not, AI is, as Taylor Searle says, a reality, so with that in mind, how do my interviewees feel about the future of creativity? ‘Excited,’ states Darji. ‘When almost anyone can make something that looks technically good, good is no longer enough; taste, emotion, originality and point of view become even more valuable. Creative technologist may become less of a job title and more of a fundamental skill every creative is expected to have’.

Spooner is feeling similarly positive. ‘I’m optimistic, he says. ‘I’ve lived through many waves of technology transforming creativity and have seen the countless ways these advancements have been assimilated by creators. So I feel curious and hopeful, but not complacent. There is work to be done, care to be taken, thinking to evolve. But there is a fantastic opportunity here – the more people that get to play with these tools, the wider range of diverging ideas we will have access to and I’m really excited to see how varied people’s responses and uses of AI are’.

Likewise Taylor Searle’s prevailing feeling about the future of creativity in an AI affected landscape is shaped by her experience of AIs’ potential users, as much as its potential uses. ‘It might sound cliché, but the future of creativity feels warm,’ she says. ‘There is an incredible community in creative computing and computational arts groups, one that looks after each other and deeply cares about our relationship with technology’.

So where does all this leave me? Am I still concerned about AI? Absolutely. But then I’m a stubborn so-and-so, so that’s on me rather than Darji, Spooner and Taylor Searle. However, their insight has left me much less dismissive of it, and I’m less fearful that it’s going to steal my job (it wouldn’t have put three ‘so’s in that previous sentence for a start). I can see its potential, particularly for making art and creativity much more democratic and accessible, and that has to be a good thing, so long as the technology is allowed to evolve that way.

To go back to the question posed in the title, is AI killing creativity? I don’t think it is, and while it’s destined to have a significant impact on its future I don’t think it ultimately will either. That’s not to say that it doesn’t pose a genuine threat, but after speaking with Darji, Spooner and Taylor Searle I am at least reassured that the power to stave off that threat and manipulate it into an opportunity for creative good is still in artists’ hands. Which is why, given AI is already here, the shape of a creative future with it needs to be discussed, rather than run from. No matter how worrying it is.