I’ve been trundling around the creative conference circuit this summer. One question has been frequently raised: “Will AI take over the creative process?”

The answer has been variations on a consistent theme, to paraphrase: “Yes, AI is changing the game, but it will never replace the fundamental human ingredients, the agency, emotion, perspective, experience and intuition necessary to be truly creative.”

Each time I hear this, I feel a sense of unease at the confidence of this response. I picture those old episodes of Tomorrow’s World predicting a future dominated by microchip implants, paper dresses, worm omelettes and the Cyberspace Riots of 2005 (all real predictions, btw), and wonder: “How can we be so sure?”.

To answer this, let’s start with a basic question: “Can AI be creative?”

This depends on the definition of creativity. AI is currently capable of generating original content, such as music, stories and images at tremendous speed with minimal human input. It can aid ideation, stimulate thought and optimise responses. Right now, it can be seen as no more than an insanely powerful tool: a mimic, a generator, an optimiser, an aggregator. But creativity lives far beyond outputs. Creativity is intention, insight and revelation. It is the expression of talent, emotion, perspective and original thought. It is the articulation of culture, the reporting of experience, the evidence of soul, the unbounding of rules. AI lacks the human spark which made creative expression the first kind of communication humans were capable of.

Put another way: AI can make creative things, but only humans can create with purpose. Yet here we are on the precipice of a kind of change which will only become truly knowable when we’ve been afforded enough time to look back and – as with the internet before – truly understand the intricate ways in which the world has been reshaped around us.

Until then, we can only guess. But it is reasonable to assume that creativity itself will also change. We can imagine, for example – in ways we can already see – that creatives will shift from makers, to curators, applying greater use of taste, judgement and storytelling to amplify AI’s outputs. Indeed, it is foreseeable that the ability to edit, prompt, synthesise and provoke may soon be more desirable than the talent to draw, the imagination to romanticise, or the capacity for prose. But what if that’s not where things stop? As AI’s advances are measured in days rather than years (Moore’s Law famously predicted computer power doubling roughly every 18-24 months; Large Language Model’s benchmarking performance on tasks like reasoning, coding and language understanding has been doubling every 3.5 months), the reasonable question remains: “Can AI become creative?”

AI’s particular flavour of creativity may always, definitionally, be routed in mimicry, rather than true creative expression. But it may achieve levels of output so convincing in quality that they become indistinguishable from human creativity (by Turing Test logic: if it looks, feels and functions like creativity, who are we to say it isn’t?).

Indeed, we might even learn to love it.

As generations grow up as AI natives, the reticence and distrust their elders innately felt may dissipate. The fear of replacement may be overridden, in its place an embracing and affinity for technologies which aid work, increase knowledge and reduce boredom…

“So what if a book isn’t authored?”, “Who cares if a picture isn’t painted?”

I’m left wondering then: “Is artificial creativity really implausible?”

Much is being said of the next technological goalpost: Agentic AI. It’s the ability for artificial intelligence to go beyond passive responses – to act with apparent autonomy, to set and pursue goals, to make decisions, take action and learn over time.

As LLMs suck up the entirety of humanity as told through the internet – the stories, pictures, music, history, memes, gifs, recipes and dad-jokes that define our earthly experience – while developing at a rate where tools that seem impressive today will be outdated within months, then perhaps its ability to reach a kind of creative singularity – where it can apply autonomous thinking to behave creatively – is not such a farfetched idea after all.

We can think of this as ‘Agentic Creativity’, where AI stops waiting to be told what to do and starts making with apparent intent. It plans, iterates and even surprises. It has both autonomy (ability to act independently) and agency (ability to act with intent). No longer a tool, but a member of the team. From instrument to collaborator (or competitor?).

Some possible scenarios for how Agentic Creativity might evolve and become evident, include:

1. AI develops creative styles AI programmes may evolve their own recognisable creative signatures, not because humans have trained it to, but because of internal, self-directed goals. An AI system generates art with unexpected colour palettes or composes music with a distinct sound that defies simple genre categorisation.

2. It independently responds to cultural changes AI systems identify cultural trends and evolves its outputs in response. It sees that certain ways of telling stories resonates strongly with audiences and so shifts towards those new forms. It understands context, effectively applies subtext and adapts its voice to better resonate across different audiences.

3. It becomes… good.

Good quality, that is… The soulless music ‘slop’, cliché-ridden aesthetics, shallow storytelling and uncanny valleys of today’s AI outputs are replaced with genuinely surprising, inspiring, beautiful and emotionally resonate results – things people knowingly and wilfully consume.

I’m certainly not advocating for a kind of post-creative (or should that be ‘post-human-creative’?) era. That world doesn’t sound like one I’d much enjoy. But I do think it’s important we consider the very real possibility that this pace of change may catch up with creators of every kind. It may well be naïve to assume that a human will always be the gatekeeper of creativity – and as much as I am anxious of the kind of world I am painting here – it may be time to consider a new question: “What happens when AI becomes creative?”

A rather meta post-script: In writing this article, AI has been an important tool in informing and ordering my views. So why does the previous sentence feel more like an admission of guilt than a statement of fact? It will require an enormous shift in perception and acceptance to embrace this tool and – particularly for a writer like me – to not feel as though I am somehow cheating the system, or indeed myself, in its assistance. In writing this, AI has helped me develop thoughts and reach conclusions I may not have otherwise realised, but it remains my hope that it is in the creative articulation of those ideas that true insight and, at least for now, uniquely human-shaped value can be achieved.