Everyone’s talking about generative AI. The word “generative” has been completely absorbed into the AI lexicon. But generative isn’t a new concept. Humans have been generative for as long as we’ve existed. And what we generate is fundamentally different from what a model produces.
It’s worth reclaiming the word.
What Generative Actually Means
Strip away the tech context and “generative” just means: creating something that didn’t exist before. That’s it. And by that definition, most of the important moments in your life are generative ones.
A conversation where you walk away thinking differently than when you sat down. Writing something that clarifies your own thoughts, not just communicates them. Teaching someone and watching the understanding land. Building something with your hands. Raising a kid. Coaching a teammate through a hard problem.
These are all generative acts. And they share something that AI-generated content doesn’t have: they change the person doing the generating.
The Bidirectional Thing
This is the part that gets overlooked. When a large language model generates text, it produces output. Tokens in, tokens out. The model isn’t changed by what it creates. It doesn’t learn from the act of creating. It doesn’t struggle, reconsider, or grow.
When a human writes, teaches, builds, or has a real conversation, the process transforms them. You sit down to write about an idea and realize halfway through that you were wrong about part of it. You explain something to a colleague and your own understanding deepens. You build a table and learn something about patience and precision that carries over into everything else.
Human generative activity is bidirectional. The output changes the world, and the process changes the creator.
Leisure, Generatively
Arthur Brooks recently made a point about leisure that connects here. He draws a distinction between passive consumption and active engagement. True leisure isn’t scrolling, bingeing, or numbing out. It’s doing something that restores you because it demands something of you. Playing music. Cooking a real meal. A conversation that goes somewhere.
The overlap is obvious. The leisure that actually recharges you is the leisure that’s generative. It asks something of you and gives something back. Not content consumed, but experience created.
Compare that to the default modern leisure: infinite scroll, algorithmic feeds, content that passes through you without leaving a mark. That’s consumption. It’s the opposite of generative.
The Scarcity Argument
Here’s where timing matters. We’re entering a period where AI-generated content is abundant and getting more abundant by the day. Text, images, video, code. The marginal cost of producing content is approaching zero.
Which means the signal value of human-generated work is going up, not down. When anyone can generate a 1,000-word blog post in seconds, the act of actually thinking through an idea, wrestling with it, and writing it yourself becomes a differentiator. Not because the output is necessarily “better” in some objective sense, but because it carries proof of thought. Proof of struggle. Proof of taste and lived experience.
People can feel the difference. Maybe not always consciously, but they can tell when something was created by someone who was changed by the creating.
So What?
This isn’t an anti-AI argument. I use AI tools constantly. They’re genuinely useful for accelerating work, exploring ideas, handling the mechanical parts of creation.
But there’s a trap in outsourcing too much of the generative process. If you let AI do all the writing, you lose the thinking that writing forces. If you let AI handle all the hard conversations, you miss the growth that comes from navigating them yourself. If you optimize away every friction, you optimize away the transformation.
The most generative thing you can do right now probably isn’t prompting a model. It’s doing the hard thing that changes you in the making. Writing the messy first draft. Having the difficult conversation. Building the thing with your hands even though it would be faster to automate.
Not because efficiency doesn’t matter. But because some things are valuable precisely because they cost you something.
That’s what generative means. It always has.