AI 让我们变得不创意了吗？Has AI Made Us Less Creative?

Plenty of people turn this question over in private and almost never hear a direct reply. Creators in one camp say AI opened doors that used to be shut to them. Creators in the other describe a slow erosion — the urge to grind through a hard problem going slack, the willingness to stare at an empty page and keep going anyway draining away. Both accounts can hold at once. What follows is the honest version.

At some point in the past two years, creativity turned into disputed territory. Designers who once put ideas down by hand now type instructions into a model. Writers who once let an idea sit for a week before producing a sentence now turn out ten versions in sixty minutes and keep whichever is least poor. Musicians release albums built on AI instruments they have no physical ability to play. Seen from outside, the volume is up. Seen from inside, plenty of creators cannot say whether what they feel is more creativity.

Those voicing the worry are not anti-technology. They are mostly the same people eagerly picking up AI tools — for the very reason that daily use shows them exactly what shifts when a sizeable chunk of the process is handed off. The objection is not that the tools exist. It is about what becomes of a skill, and of a creative self-image, once regular practice stops.

01The Short Answer, Up Front

Creativity is not vanishing because of AI. What is changing is which creative muscles do the work — and that is not trivial. The worry rests on real ground: cognitive science has long held that unpractised skills decay, and creative thinking counts as a skill. Let AI take over the decision-making portions of a task on a regular basis, and the human at the keyboard makes fewer such decisions. Something shifts over time as a result.

Meanwhile, genuine creators are producing work that is interesting and original with these tools — work that simply would not exist otherwise. Both statements hold at the same time, which is exactly why framing the issue as "is AI killing creativity" generates heat rather than light. A better question: inside your creative process, which steps has AI taken over, and which are still yours?

02Settle What Creativity Means Before Arguing About It

Debates about AI and creativity often spin in place because the participants never agree on what the word covers. Take creativity to mean producing something unprecedented, and AI plainly qualifies. Take it to mean voicing something true about lived human experience, and crediting a model whose training came from human writing and pictures with that becomes far harder. Whether AI is creative therefore hinges on which sense you have in mind.

For this article, adopt the definition most cognitive scientists and working artists use: the ability to generate original ideas, draw connections nobody expected, and express something carrying a real personal viewpoint or history. Mere newness is not enough — that is just randomness. Creativity is novelty with a purpose, formed by taste, by judgement, and by having a reason to speak.

By that standard, AI does not create. It emits outputs that are statistically probable given the patterns in its training data. Those outputs can surprise you and sometimes are beautiful, but they lack purpose and carry no viewpoint, so they are not creative in the sense just described. The question that matters here is not whether AI is creative; it is whether using AI changes whether people are.

03The Argument for Yes: When AI Really Does Cut Creativity

It is worth taking the uncomfortable half of this argument at full strength. Certain ways of using AI visibly lower creative engagement, and pretending they do not helps nobody.

What actually alarms researchers in cognitive science is the steady handover of the effortful, struggle-heavy, decision-laden portion of the work. The link between how hard creative production is and how creative skill develops is well documented. Wrestling with a problem — even when it goes nowhere — builds something. Someone who grinds out ten poor paragraphs to reach one good one has been practising judgement. Someone who has the model produce fifty paragraphs and keeps the least poor one has practised something else entirely.

04The Argument for No: When AI Really Does Widen Creative Range

The opposing claim is equally real. Right now there are creators making things that could not have been made before — and the reason is not that AI drops the standard, but that it removes obstacles unrelated to creative talent: technical skill, hours, money.

A novelist with a visual imagination but no drawing hand can now illustrate the world they invented. A composer who works on paper can hear a full orchestration without paying an orchestra. A designer with animation concepts but no coding ability can build a moving prototype for the first time. In every case the creativity — the idea, the vision, the intention — belongs entirely to the human. What AI covers is the technical execution gap, not the creative calls.

History supplies the strongest counter to "AI makes us less creative." Every significant productivity tool was said to be degrading the very skill it helped with. Word processing would ruin writers; calculators would ruin mathematicians; photography would finish off painting. The worry was sincere each time, and each time the reality was messier: the tools rearranged which skills counted and let practitioners operate one layer further from the details. Something of the same kind is probably underway.

05The Blank Page Problem, and Why It Deserves Attention

Saying that AI creative tools "solve the blank page problem" has become a favourite refrain. No more sitting there with no notion of what to make — generate something first, then respond to it. That is accurate, and plenty of people find it genuinely helpful. Still, honesty requires asking what the blank page had been doing for you.

An empty page is unpleasant to face. It is also the site of a particular sort of creative work: the stretch of time when an idea has no shape yet and your mind sifts connections and options with no outside input. The most consequential creative thinking — the kind that yields work nobody saw coming — often occurs in that awkward interval before anything is written. Remove that interval every single time and it is far from obvious that nothing gets lost.

Being able to face an empty page and force your way to something is itself a practised ability. Writers who do it build a particular strength: generating from their own resources, trusting their instincts, and putting up with the uncertainty of not yet having an answer. Whether that strength counts for anything in a world where the gap can be filled the instant it appears is genuinely unsettled — but it deserves careful thought before anyone writes it off.

06Skill Atrophy: Genuine Risk or Overblown Fear?

Cognitive offloading — handing mental work to an outside tool that you used to do in your head — is a well-established phenomenon. Constant GPS use has measurably weakened spatial memory. Spellcheckers have altered how people relate to spelling. Nobody would call such losses catastrophic, but they are losses. The open question for AI is whether creative cognition behaves the same way.

Honestly, the long-term data does not exist yet. Creative AI tools at this level of adoption are brand new. Given what is known about skill and practice, the fear that routinely outsourcing creative decisions erodes the ability to make them alone is a reasonable one — but it remains a reasoned concern rather than a demonstrated result. Claim that the evidence clearly settles the matter either way and you are running ahead of the research.

07What AI Genuinely Cannot Do

One ingredient makes creative work worth the making, and AI does not possess it: a reason to speak. Work that lasts tends to come from someone with something they urgently must get out — a viewpoint that feels like nobody else's, an experience that has to be worked through, an observation about the world that will not stop nagging them. AI has no experience to draw on, no urgency, no inner life pressing for expression.

That is why AI-made creative output, even when the craft is dazzling, usually lands as faintly hollow. It has statistical competence. It has absorbed the patterns of what succeeds. What it lacks is any stake in the outcome. The unmistakable voice that makes a piece worth reading, viewing, or hearing originates in a particular person who has lived, thought, and felt particular things. Prompt engineering cannot manufacture it.

Practically speaking, that leaves one set of skills worth investing in: the ones no AI can take over. Holding a perspective nobody else holds. Building real taste by consuming creative work widely and critically. Acquiring the judgement to tell what is good apart from what merely resembles good. Coming up with a motive for making things that belongs to nobody but you. These do not decay because of AI — AI simply does not operate there. Whether the people best placed to develop them are actually spending the time is a separate matter, and a far more personal one.

08Using AI Without Losing Yourself

Most creators are not choosing between refusing these tools and swallowing them whole. What they want is a working arrangement in which the work stays theirs. A handful of principles that travel well across disciplines:

• Decide before you prompt. Be clear about the outcome you are after before you ask the model for help getting there. Letting AI supply the idea while you pick from its menu is a different mental exercise from using it to carry out an idea you arrived at yourself.
• Keep one practice that runs without AI. Write by hand now and then. Sketch with nothing generating alongside you. Draft the first pass unaided. The reason is not that AI is harmful — it is that the unaided version works a capacity the aided version leaves idle.
• Build taste on purpose. Read across a wide range, study work you admire, and form real opinions about what is good and the reasons why. Taste is the line between creative judgement and aesthetic guesswork, and no model can grow it on your behalf.
• Be straight with yourself about the model's share. If AI produced both the concept and the execution while you selected and hit publish, what creative work did you actually do? Sometimes the honest answer is curation and direction — a legitimate creative role. Knowing which one applies still matters for how you develop.

How AI is reshaping creative work as a profession — which roles it touches and in which direction — ties straight into the wider employment picture across creative industries. Our analysis of what AI means for content writing jobs in 2026 digs into the economic and career side of the same question, and reads well alongside this piece.

09The Risk That Gets Discussed Too Little

One creative risk gets far less airtime than skill atrophy, and in my view it is the more pressing of the two: everything starts to look alike. Put many people on the same tools, trained on the same corpus, prompting in broadly the same way, and the results drift toward one another. Anyone who has looked at enough AI images can already spot the house style. The sentence shapes that AI writing tools reach for are becoming a tell. AI-composed tracks carry recognisable fingerprints.

None of that is aimed at individual creators who use the tools well. The point is structural: route creative production through a handful of very capable systems and the variety of output across a culture narrows. The long tail of odd, uncategorisable, deeply strange human work — the material that emerges from one particular mind at one particular moment — now competes on different terms, because polished and technically assured AI output appears instantly and costs almost nothing.

10Questions Readers Ask

11Where This Leaves Us

"Is AI making us less creative" honestly resolves to: it depends — and that is not a dodge. It depends on which stages of the process you gave away, how regularly you practise creating unaided, whether you are building real taste and a real viewpoint, and whether your output still contains something no one else would have produced. Every one of those is a question about personal habit, not about whether the tools exist.

One thing does look clear: the creative capacities most worth defending are precisely the ones AI cannot reproduce. Accumulated experience, which is what gives you something worth saying. Taste, which separates the genuinely good from the merely acceptable. And the patience to remain in the difficult place long enough that something original can surface. AI tools do not threaten any of those. What they do demand is deliberate cultivation, given a landscape where the effortless route sits one prompt away. Whether anyone chooses that cultivation is — fittingly — a very human creative decision.