Algorithmic Gravity

Art and Algorithmic Gravity

There’s a voice that arrives quietly, after enough time on platforms.

Not loud. Not commanding. Just a second thought that can trail behind a new idea: but will it perform?

It doesn’t announce itself. It accumulates. Years of analytics, retention graphs, click behavior, not forcing anything, just suggesting. Rewarding. Shaping, the way water shapes stone: invisibly, persistently, over a long time.

I’ve been putting creative work into the world online for roughly twenty years. Long enough to remember when platforms felt less like casinos and more like open fields. You made something, posted it, and maybe it found people. The process was imperfect but comparatively neutral. Depth could still find its audience. Patience was still a viable strategy.

That has changed. Not through any dramatic rupture, but through something subtler and harder to name.

I’ve started calling it algorithmic gravity.

Gravity doesn’t command. It doesn’t threaten or censor or forbid. It simply exerts a pull so constant, so precisely calibrated to human psychology, that over time behavior bends around it without anyone deciding to bend. Artists shorten videos because people click away. Simplify their process because the numbers say it performs better. Return to the same style, the same format, the same emotional register, because repetition is what the system quietly, persistently rewards.

None of it feels like compromise in the moment. That’s what makes it dangerous.

The question changes. Slowly, almost imperceptibly, the center of gravity shifts from what do I want to make to what performs and that shift, small as it sounds, changes everything. Performance becomes a kind of creative prison with invisible walls. The systems that promise visibility reward consistency over evolution, recipe over exploration, familiarity over the kind of risk that produces something genuinely new.

And yet.

Meaningful work almost never emerges from optimization. Deep work is inefficient. Experimentation is inefficient. Evolution is inefficient. Some of the most important creative breakthroughs happen when an artist disappears into something difficult for months, sometimes years, with no guarantee it will resonate, no metric to validate the direction, no algorithm to consult. From the outside, truly important work often looks like bad strategy. It confuses its audience. It underperforms. It refuses easy categorization.

But it matters. Often more than anything that scored.

The pull has grown stronger and more psychologically precise over two decades. The temptation to simplify, to repeat, to favor what lands quickly over what unfolds slowly, to become increasingly efficient at producing variations of what already worked, is not a personal failing. It’s an engineered outcome. These systems are very good at what they do.

Which is precisely why resisting them requires something more than willpower. It requires a different set of questions entirely. Not what performs but what matters. Not what does the audience want more of but where does the work actually need to go.

The things that genuinely serve art, slowing down, going deeper, risking failure, disappearing for a while, making something that doesn’t fit the recipe, are precisely the things the algorithm punishes most. There is no metric for the two years spent on something that almost broke you. There is no engagement score for the work that arrived, finally, only after you stopped asking whether it would perform.

That work exists outside gravity’s reach.

Which is, I think, exactly where it needs to be made.