NOT long ago, artificial intelligence made headlines for what many described as a creative breakthrough. OpenAI’s GPT-4 ranked in the top one per cent on standardised creative thinking assessments — those same tests used to measure human originality and problem-solving. Around the same time, an Artifical Intelligence generated song that mimicked the voices of Drake and The Weeknd went viral, collecting millions of streams before legal questions forced it offline.
These moments were technically impressive. But they also forced a question that’s no longer hypothetical: can machines genuinely replicate, or even replace, human creativity?
The short answer is complicated. Yes, AI can generate content that appears creative. It can sketch, compose, write, remix. It can mimic styles, recognise patterns and combine information in ways that sometimes surprise us. But the longer answer, and the more important one, suggests that there remains something essentially human at the core of creativity — something that machines, at least for now, cannot reach.
Over the past few years, AI’s creative capabilities have expanded with astonishing speed. In 2025, we are now seeing machine-generated paintings, stories and music that, on a surface level, can stand alongside human works. AI has found a place as a collaborator across industries — from advertising to architecture, from fashion to film. Large language models don’t just predict likely words; they are now trained to produce persuasive copy, pitch business ideas and draft poetic verses. Visual models churn out images that make it to magazine covers and gallery walls.
This growth is partly tied to how creativity is measured. Standardised tests like the Torrance Tests of Creative Thinking focus on traits such as fluency, flexibility and originality. These are tasks with prompts and goals — contexts in which machines excel. Given a structured environment, AI can shine.
But that is also the problem. Real-world creativity is not a prompt. It arises from memory, contradiction, discomfort, boredom and beauty. It emerges not just from data, but from perspective. Two people may respond to the same image or question in wildly different ways depending on who they are, what they have lived through, or how they feel. AI, by its very nature, lacks that kind of interior life. It does not grow up in a remote village, or lose a parent, or fall in love on a rainy afternoon. It does not dream, or fear, or grieve. It processes; it does not experience.
And this absence becomes visible in the work. AI-generated stories, however well-structured, often lack the emotional depth or unpredictability of human writing. They follow patterns. They converge toward coherence, not challenge. The results can be impressive — neat, clever, even charming — but rarely moving.
There is a reason why human-made art continues to hold us. We’re not only looking for novelty or beauty; we’re looking for truth. We want to feel someone else’s thoughts and emotions coming through the work. We want connection. And connection cannot be reverse-engineered through code.
As one researcher put it, people create because they feel something. They make art to mourn a loss, to preserve a memory, to make sense of their pain or joy. Machines, regardless of how advanced, do not feel. They do not make because they must. They make because they are asked to. That is not a failure — but it is a limit.
Still, framing this as a rivalry between humans and machines misses the point. The more productive conversation now is about collaboration. Artists and technologists are increasingly asking: what can humans and AI do together that neither could manage alone?
The Max Planck Institute has studied creative partnerships between people and machines, showing how AI can expand the artist’s toolkit — offering suggestions, generating prompts, stretching boundaries. Musicians, illustrators, even choreographers are using AI not to replace their own voice, but to enrich it. Like discovering a new instrument, or a colour never seen before.
This kind of collaboration holds promise. But it also raises urgent questions. UNESCO and other global bodies have warned of the risks: the need for transparency around AI-generated work, the ethical use of training data and fair compensation for human creators whose labour trains these models. Without careful regulation, we risk flooding cultural spaces with algorithmic content that flattens difference and drowns out marginal voices.
And there is the wider economic context. The World Economic Forum estimates that AI and automation could displace 85 million jobs by 2025, even while creating 97 million new ones. But numbers don’t tell the full story. If creative tasks — writing, designing, composing — are increasingly handled by machines, what happens to the people behind those tasks?
This doesn’t mean the death of creativity. But it may require us to redefine what counts. What survives the wave of automation might be the most deeply human aspects of creation: storytelling, emotion, intent. The part of writing that doesn’t just communicate, but means something.
We may soon see a sharper divide between routine creativity — like drafting corporate slogans or social media captions — and generative creativity that asks bigger questions, takes risks, explores meaning. The first can be automated. The second is harder to teach and harder to imitate.
That shift should reshape how we teach and value creativity. Education systems must foster not only technical competence, but also curiosity, empathy, narrative thinking. The ability to see connections across disciplines. To ask strange questions. To reflect. To care.
We should also begin to appreciate creativity not simply as content production, but as a reflection of identity. In a world of infinite AI output, originality won’t just be a matter of novelty, it will be a matter of voice.
In the end, the question is not whether AI can be creative. It clearly can, in the narrow sense. The real question is what we want from creativity. Is it efficiency? Surprise? Expression? Empathy? Depending on our answer, the future could take very different shapes.
What is becoming clear, though, is that creativity — at its richest — is not going anywhere. The tools may change. The collaborators may be digital. But meaning, vulnerability, and vision still rest with us.
Machines may generate. But only people can mean.
Ìý
Md Ibrahim Khalilullah is a writer focused on the intersection of technology, culture and society.
