THE Algorithm Didn't Kill Creativity: It Forced Us to Grade It

​Not long after generative artificial intelligence entered the public imagination, an AI-generated song mimicking Drake and The Weeknd racked up millions of streams before platforms pulled it for alleged copyright infraction. The argument that followed was about ownership – who made it, who gets paid, who gets sued.
​
But an equally important question was voiced much quieter, and has yet to be answered: 

Why did it sound so good to us in the first place?

​Because the algorithm knew the formula.

AI replicated what had already worked and served it back as novelty. In doing so, it revealed how easily we mistake pattern recognition for creativity itself. The result wasn’t a creative leap; it was perfectly average – the antithesis of creativity itself – and unfortunately, exactly the sort of “creativity” our systems are training us to reward.

Now Comes AI
I earned a degree in creative writing long before anyone could type a sentence into a chatbot. But artificial intelligence changes things. Now you can simply ask ChatGPT to “make this more creative.” It’s a strange command, like outsourcing your epiphany. Creativity used to be something we nurtured; now it’s something we seek to optimize.

Economists have a name for what happens when optimization takes over: Goodhart’s Law. “When a measure becomes a target, it stops being a good measure.” That applies to non-economic measures as well.

Pay teachers by test scores and they’ll teach to the test; reward hospitals for shorter wait times and some patients are discharged too soon. Apply the same logic to art and the results are familiar. Once we start to measure creativity through clicks and “likes,” we start designing to the metric.

Now platforms report, in real time, how every frame, lyric or sentence performs. The video’s hook lands in the first three seconds. The headline is written for the search engine. The familiar chorus hits before the listener can skip the track.
We don’t just respond to metrics; we anticipate them.

The Predictive Dilemma
Artificial intelligence perfects this unfortunate reality. AI isn’t creative; it’s predictive. It learns the patterns we’ve already rewarded, the chord progressions or color palettes or turns-of-phrase that keep us scrolling. It is creativity’s autocorrect, making everything recognizable, nothing surprising.

But the better the AI model gets at predicting what we like, the narrower “what we like” becomes. Soon the loop closes: AI imitates our tastes; we imitate AI’s outputs; the next model trains on those imitations.

Innovation morphs into convergence, and what disappears in that convergence is discovery beyond the ordinary. Algorithms are built to reward what is common, to serve us more of what we already like. In this feedback loop, experimentation becomes a liability. The strange gets smoothed out, the risky gets filtered away, and discovery – the engine of originality – slowly disappears.

What We Lose
There’s a human cost to this. When creativity becomes something to optimize, we live according to algorithms that are built to deliver what’s common. Students chase the accepted rubric instead of a potentially unique vision. Employees brainstorm inside the boundaries of what’s been rewarded before. Artists second guess their creative instincts.

In my classes, I’m asked for examples of A-level work before a project begins – they want proof before process. It’s an understandable impulse in a system that equates certainty with virtue, but it leaves little room for imagination. Teresa Amabile calls this the intrinsic motivation principle: creativity thrives under curiosity and autonomy; it withers under surveillance and reward.

Beyond Measurement
So, what would it take to rescue creativity from the grip of measurement?

Not better metrics, fewer of them. Or at least a different relationship to them. Creativity needs room to stumble, contradict itself, and – importantly – discover what we didn’t even know we were looking for.

In education, that means grading process alongside product: elevating reflection logs, drafts, risk-taking, structured discussion into the evaluative dynamic.

In organizations, it means rewarding experiments that fail interestingly and treating iteration as a sign of vitality, not waste.
For individuals, it is creating for meaning rather than measurement – reclaiming the joy of creativity for its own sake.

Trust Rules
In the end, we don’t need better dashboards for creativity; we need better trust. Constructive critiques instead of grades, dialogue instead of rankings, patience instead of measuring immediate performance.

The dashboard isn’t the canvas. The canvas is the messy, unpredictable, deeply human space where we strive to understand and find meaning that can’t be optimized. If we can find our way back there – beyond the metrics and mirrors – we might help our students come back around to what creativity is all about in the first place.
​
Maybe the only way forward is to flip the question. Instead of asking whether creativity can perform, ask whether it can still transform – our work, our lives, our culture.