What AI Is Doing to Our Definition of Originality

There was a time when the ability to make something was itself a significant part of the achievement.

If you wanted to paint a convincing portrait, you needed years of practice. If you wanted to compose music, you needed some command of an instrument, notation, harmony, or arrangement. If you wanted to publish a book, you needed to write the manuscript, edit it, find a publisher or learn how to publish it yourself, and navigate an entire infrastructure between the first sentence and the finished object.

The technical difficulty was part of the creative economy.

Generative AI is changing that.

A person can now describe an image and receive dozens of visual possibilities. A writer can ask for alternative structures for an argument. A musician can explore arrangements that would once have required considerable technical knowledge. A designer can generate variations before deciding which direction deserves further attention.

This does not mean that creativity has become effortless. It means something more specific has happened:

The cost of producing possibilities has fallen dramatically.

And once possibilities become cheap, our definition of originality has to change.

Novelty Was Never the Whole Story

We often use original to mean new.

That seems reasonable until we examine what human creativity has actually looked like throughout history.

Most creative work is built from something that already existed.

A novelist inherits language, narrative structures, genres, and cultural memories. A painter inherits visual traditions. A musician inherits rhythm, harmony, instruments, and forms. A scientist builds on previous discoveries. A filmmaker borrows conventions and then alters them.

Human beings are extraordinary recombiners.

That means the arrival of AI does not introduce imitation into a previously pure world of human originality. Humans were already drawing from enormous cultural archives.

The difference is that generative systems can perform certain forms of recombination at a scale and speed that individual humans cannot.

Research published in Nature Human Behaviour compared 9,198 humans with more than 215,000 large-language-model observations on a divergent-creativity task. The study found that humans and LLMs have overlapping creative capabilities, while humans retained an advantage in the highest levels of creative performance. Read the study.

That finding matters because it complicates the easiest story about AI.

The machine is not simply incapable of creativity.

Nor has it demonstrated that human creativity has become obsolete.

Instead, we are discovering that creativity contains multiple components, and humans and machines may be good at different parts of the process.

That distinction is going to matter enormously.

The Real Shift Is From Making to Choosing

Imagine that you have a machine capable of generating ten thousand paintings.

The machine can produce them.

But which one should exist?

That question remains.

You might choose the most technically impressive painting. Someone else might choose the strangest one. Another person might reject all ten thousand because none expresses what they were actually trying to say.

The generation is automated.

The judgment isn't.

This is why the future of creativity may belong disproportionately to people with unusually developed taste.

Taste is sometimes treated as a superficial aesthetic preference. In creative work, it is much more consequential.

Taste is the ability to recognize the difference between something that merely works and something that matters.

It tells an editor which paragraph should survive.

It tells a filmmaker which frame is worth keeping.

It tells a musician when a technically perfect arrangement has lost the feeling of the song.

It tells an artist when an imperfect mark is more important than a polished one.

And it tells a writer when an apparently clever sentence is actually saying nothing.

AI can give us more options.

That makes choosing among those options more important.

The paradox is that abundance can increase the value of discernment.

The Creator May Become the Person Who Knows What to Reject

For centuries, creative education largely emphasized learning how to make things.

The future may place much greater emphasis on learning how to reject them.

This sounds strange until you consider what happens when generation becomes nearly unlimited.

If you can produce fifty headlines in thirty seconds, the problem is no longer generating headlines.

It is knowing which headline deserves to introduce the argument.

If you can produce a hundred visual concepts, the difficult part is recognizing the one that actually communicates something.

If you can generate twenty possible endings to a novel, the challenge becomes understanding which ending belongs to the story.

Creation therefore begins to move upstream.

The creative act is increasingly located in the questions, constraints, selections, revisions, and reasons that surround generation.

That is a profound change.

It means that someone who knows exactly what they are trying to make may have a greater advantage than someone who simply knows how to generate more.

But This Is Where the Argument Gets Uncomfortable

It would be convenient to say that AI generates and humans create.

Reality is more complicated.

A person using AI may ask questions, compare outputs, reject suggestions, combine ideas, restructure material, introduce personal knowledge, revise language, verify information, and make hundreds of decisions throughout the process.

At what point does the machine become a collaborator?

At what point is the human merely selecting?

At what point does selection itself become authorship?

These are not abstract questions.

They are already becoming questions for publishing, science, law, art, and education.

Nature has recently examined AI-mediated authorship through the related question of responsibility: a human contributor can explain their contribution, respond to criticism, and be held accountable for the work, while an AI system cannot take responsibility for what it produces. Read the discussion.

That suggests that authorship may increasingly have to be understood through contribution and responsibility, rather than through a simple binary of human versus machine.

The interesting question is therefore not whether AI touched the work.

It is:

Who made the consequential decisions?

Originality and Authorship Are Not the Same Question

This distinction is essential.

Something can be aesthetically original without being legally protected by copyright.

Something can be heavily assisted by AI while containing substantial human creative authorship.

Something can also be technically made by a human while being derivative, formulaic, or almost entirely imitative.

The law is therefore not simply asking the philosophical question, “Is this original?”

The U.S. Copyright Office's current position is that generative-AI outputs can receive copyright protection when a human author determines sufficient expressive elements. Human-authored material perceptible in an output, or creative human arrangement and modification of AI-generated material, can qualify; merely providing prompts is not enough. Using AI as an assistive tool does not automatically prevent copyrightability. Read the U.S. Copyright Office report.

That is an important distinction because public discussion often collapses several different ideas into one word.

Originality.

But originality in aesthetics, originality in culture, originality in authorship, and originality in copyright are not identical concepts.

The AI era is forcing us to separate them.

The Part AI Cannot Simply Manufacture: A Life

There is another dimension of creative work that is harder to quantify.

A human being makes things from somewhere.

A particular childhood.

A particular body.

A particular culture.

A particular set of losses.

A particular education.

A particular obsession.

A particular relationship with death, money, love, beauty, failure, religion, politics, work, or time.

Those experiences influence what a person notices.

They determine which questions become important.

They affect what feels obvious and what feels strange.

They determine what a person cannot stop thinking about.

This is one reason two people can have access to exactly the same AI model and produce profoundly different work.

The model may be shared.

The lives are not.

That distinction becomes particularly important when we talk about authenticity.

Authenticity is not simply another word for originality.

A work can be original without revealing much about its creator. Conversely, a work can use familiar ideas and still feel deeply authentic because the creator has brought an unmistakably personal relationship to them.

AI can generate an image of grief.

It can generate an essay about loneliness.

It can generate a poem about losing someone.

But the existence of those outputs does not mean the system has experienced grief, loneliness, or loss.

The human question is therefore not only:

Can the machine produce something that resembles an expression of experience?

It is also:

What relationship does the maker have to what is being expressed?

That question may become increasingly important as generated material becomes more convincing.

AI Could Actually Make Some Humans More Creative

There is another side to this conversation that deserves equal attention.

AI can expand creativity.

It can help people test ideas quickly. It can expose them to possibilities they would not have generated alone. It can help a person move between disciplines. It can lower technical barriers that previously prevented people from expressing an idea.

A person who has always imagined a visual world but cannot draw can now begin exploring it.

A writer can test structural possibilities.

A designer can prototype.

A musician can experiment.

A researcher can interrogate an argument from multiple angles.

Used intelligently, AI can become a kind of creative laboratory.

That does not make the resulting work automatically original.

It does, however, challenge the assumption that creativity is valuable only when every step of the production process is performed manually.

Photography already complicated that assumption.

Digital editing complicated it again.

Sampling transformed music.

Desktop publishing transformed writing.

Computer graphics transformed visual art.

Creative tools have repeatedly changed the boundary between conception and execution.

AI is doing this at a much larger scale.

The important question is what happens to the human being after the tool becomes more powerful.

The Danger Is Not That Everyone Will Become Creative

The more interesting danger is that everyone may begin producing things that look creative.

Those are different outcomes.

Generative systems are exceptionally good at producing plausible forms.

That creates an enormous opportunity.

It also creates a problem.

When the internet becomes saturated with competent images, competent essays, competent advertisements, competent music, and competent videos, competence itself becomes less distinguishing.

We may enter an era of aesthetic abundance in which the average quality of output rises while the amount of material worth remembering does not rise at the same rate.

This is already a concern in research on AI-assisted creativity. The question is not only whether AI can increase individual creative output, but whether widespread reliance on similar systems can push culture toward convergence.

If everyone has access to similar models trained on overlapping cultural material, there is a possibility that creative ecosystems become increasingly polished while becoming less differentiated.

The work can look excellent.

And still feel strangely familiar.

That is the danger of generic excellence.

The Most Original Person May Be the One Who Has the Strongest Question

This changes how we should think about creative education.

We have traditionally asked people:

Can you draw?

Can you write?

Can you code?

Can you compose?

Can you design?

Those questions still matter.

But increasingly we should ask:

What are you curious about?

What problem can you see that other people have overlooked?

What do you believe that you can defend?

What experience has changed how you see the world?

What contradiction keeps bothering you?

What do you want to understand?

What are you willing to spend five years investigating?

Those questions point toward a deeper form of originality.

Because when production becomes abundant, the question becomes a differentiator.

The machine can give you answers.

But the quality of the question determines what kind of answers become possible.

Originality May Move From the Output Into the Process

This may be the most important shift of all.

We have often evaluated originality by looking at the finished object.

Is the painting unusual?

Is the book distinctive?

Is the photograph unlike other photographs?

Is the design fresh?

But AI makes the finished object increasingly difficult to use as the only evidence of creative authorship.

Two visually similar images could have radically different creative histories.

One might be generated from a generic prompt and accepted immediately.

Another might emerge after months of research, hundreds of discarded concepts, personal photographs, sketches, experiments, and deliberate revisions.

The final images might look superficially similar.

The creative processes are completely different.

This means that creative provenance may become more important.

We may care more about how something was made, what decisions shaped it, what sources informed it, who revised it, and who took responsibility for the final result.

The history of the object becomes part of our understanding of the object.

What Happens to Artists When Execution Becomes Cheap?

Artists will still need technique.

But technique may become less synonymous with manual execution.

The artist of the future may be part painter, part director, part editor, part researcher, and part curator.

The writer may become increasingly concerned with conceptual architecture, evidence, voice, structure, and selection.

The filmmaker may direct both human performers and generative systems.

The designer may spend more time establishing systems of judgment than manually producing every variation.

This does not make craft irrelevant.

It changes what craft includes.

Knowing what to ask can become a craft.

Knowing what to discard can become a craft.

Knowing when an AI suggestion has made the work worse can become a craft.

Knowing when to stop generating and start thinking can become a craft.

And Then There Is the Question of Meaning

This is where the discussion eventually becomes philosophical.

Suppose AI becomes extraordinarily good at producing beautiful things.

Suppose it can write moving stories, compose extraordinary music, and create images that people genuinely love.

What happens then?

We cannot answer that question merely by saying humans will still be more creative.

Perhaps humans will not always be.

Perhaps machines will eventually outperform humans on many conventional measures of creativity.

That possibility does not eliminate the human question.

It makes the question more important.

Because creativity is not valuable only because the creator performed a difficult task.

We value art because of what it does in human life.

A song can become attached to a marriage.

A painting can become associated with a death.

A novel can change the way someone understands their childhood.

A photograph can become evidence of a person who is no longer alive.

A book can give someone language for an experience they could not previously explain.

Meaning is not contained entirely in the production mechanism.

It emerges through relationships between objects, creators, and audiences.

AI changes the production mechanism.

It does not eliminate those relationships.

The Future May Be Less About Human Versus Machine

The most interesting creative future is unlikely to be a clean competition between humans and AI.

It will be a negotiation.

Humans will delegate some forms of execution.

Machines will generate possibilities.

Humans will select.

Machines will propose.

Humans will reject.

Machines will surprise.

Humans will decide whether the surprise belongs.

And throughout that process, the central creative question will remain:

What are we trying to make, and why?

That question is harder than generating an answer.

It requires a worldview.

It requires values.

It requires attention.

It requires taste.

It requires the willingness to care about one possibility more than another.

And sometimes it requires the courage to decide that nothing generated so far is worth keeping.

The Rarest Creative Resource May Be Attention

There is a final paradox hiding inside all of this.

AI is making production abundant.

But human attention remains limited.

That means the future will not necessarily reward whoever can create the most.

It may reward whoever can create something that earns sustained attention.

This changes the economics of culture.

When there are millions of generated images, one more image is almost meaningless.

When there are millions of essays, one more essay has to earn its existence.

When there are millions of songs, one more song must give someone a reason to listen.

The question shifts from:

“Can this be made?”

to:

“Why should anyone spend their finite attention on it?”

That may be the most brutal test creativity has ever faced.

And perhaps it is also a useful one.

So, What Is Originality Now?

Originality was never simply the ability to produce something nobody had ever seen before.

Human culture does not work that way.

Originality has always involved relationships between existing materials and new perspectives.

What AI changes is the scale.

The machine can now participate in recombination at extraordinary speed.

It can make possibilities abundant.

It can lower technical barriers.

It can surprise us.

It can even contribute to genuinely creative processes.

But that makes the human contribution more difficult to locate, not less important.

The human contribution may increasingly live in the things that cannot be outsourced without changing the nature of the work:

the question, the intention, the selection, the context, the judgment, the lived experience, the responsibility, and the reason for making the thing in the first place.

The future of originality may therefore look very different from the originality we were taught to admire.

It may have less to do with making something from nothing.

It may have more to do with seeing something differently.

And perhaps that is what originality was all along.

Frequently Asked Questions

Is AI making human creativity less original?

AI makes it easier to generate novel material, but novelty and originality are not identical. The larger effect may be that production becomes less scarce while judgment, perspective, selection, and meaning become more important.

Can AI create original art?

AI can generate outputs that are novel and aesthetically distinctive, and research shows that AI systems can perform strongly on some measures of divergent creativity. Whether an output should be considered authored or originally created by the AI is a separate philosophical and legal question.

What is the difference between originality and novelty?

Novelty concerns whether something is new or unusual. Originality can involve novelty, but also encompasses distinctive perspective, meaningful creative decisions, context, and the relationship between the creator and the work.

Does using AI mean that something is no longer original?

No. Using AI does not automatically eliminate human originality. What matters is how the system is used and where the meaningful creative contribution comes from. A human can use AI as part of a larger process involving research, conceptual development, selection, modification, and creative judgment.

Can AI replace artists and writers?

AI can replace or reduce the need for some forms of creative production and technical execution. That does not necessarily mean it replaces the broader role of artists and writers, whose work also involves interpretation, judgment, perspective, cultural meaning, and decisions about what deserves to exist.

Why might taste become more important because of AI?

When AI can generate enormous numbers of possibilities, the scarce skill becomes recognizing which possibilities are valuable. Taste helps a creator distinguish between what is merely competent, attractive, or unusual and what genuinely serves the intended work.

Is human experience what makes art authentic?

Human experience is one important source of authenticity, particularly when a work expresses a creator's particular history, perspective, and relationship to its subject. Authenticity and originality are related but distinct concepts.

Can AI make humans more creative?

Yes. AI can reduce technical barriers, generate possibilities, provide unexpected combinations, and allow creators to experiment rapidly. The effect depends heavily on how the person uses the system and whether AI expands their thinking or simply replaces it.

What does AI mean for the future of originality?

Originality may increasingly move away from the sheer ability to produce something new and toward the ability to formulate important questions, make distinctive choices, exercise judgment, and create work with a meaningful reason behind it.

Can AI-generated work be copyrighted?

Under the current U.S. Copyright Office position, AI-generated material can be part of a copyrightable work when a human author contributes sufficient expressive authorship. Merely supplying prompts is not enough on its own, while human-authored material, creative arrangement, or modification can contribute to copyrightability. Read the U.S. Copyright Office's guidance.

What will matter most to artists in an AI-generated world?

Technical skill will continue to matter, but conceptual thinking, research, taste, editing, discernment, originality of perspective, cultural knowledge, and the ability to create work that gives people a reason to pay attention may become increasingly valuable.

If AI can generate almost anything, what is left for humans to create?

The human task may increasingly be deciding what should be created and why. Humans bring questions, values, experiences, relationships, intentions, and consequences to creative decisions. AI can expand the field of possibilities; humans still have to decide which possibilities deserve a place in the world.

Further Reading

• U.S. Copyright Office — Copyright and Artificial Intelligence — current guidance on human authorship and copyrightability of generative-AI outputs.
• Nature Human Behaviour — A large-scale comparison of divergent creativity in humans and large language models — large-scale empirical comparison of human and LLM creativity.
• Nature — Who is responsible when AI helps to write science? — discussion of authorship, accountability, and AI-mediated intellectual work.

---

Written by Nikita Datar.