Ira Glass has this quote I’ve loved for years about the brutal gap between taste and skill. You get into creative work because you know what good looks like. Then you start making things and discover, with horrifying clarity, that you cannot yet make the thing you can recognize.

His answer is basically: make a lot of bad work. Keep going until your ability starts catching up with your taste.

AI Enters the Gap

Now AI has entered that gap—not only as another source of mediocre work, but as a tool beginners can use to make their work look more finished before their underlying skill has caught up.

LinkedIn recently got a lot of attention for doing something that looks like a contradiction. It offers AI tools to help people draft posts and improve their profiles while also encouraging members to flag low-quality content commonly described as “AI slop.”

And what is AI slop if not low-effort, low-quality work? Images, words, whatever doesn’t rise to our level of taste.

So now we have a mechanism that can help close the visible gap between taste and skill. At the same time, we’re all on the hunt for slop. The algorithms will get better at detecting and excluding it.

But What About Human Slop?

But what about human slop?

Slop and bad work aren’t quite the same thing. Slop is work with very little behind it. Bad work can be someone reaching for something they don’t yet know how to make.

Sure, some of it is just some overworked guy in Sales earnestly, if lazily, trying to get traction by posting the ol’ classic “What my 2008 Corolla taught me about B2B Sales” nonsense, somehow oblivious to the trope it has become.

But some of it is also likely to be the work of some kid fresh out of high school or college just trying to find their voice. And now I wonder if they are more likely to be shouted into a silent void as the algorithm hides that work from the feeds of their audience.

LinkedIn explicitly says it is searching out generic content rather than amateurish content today. But what if those systems struggle to discern the difference as they, and the content they’re reviewing, continue to evolve?

Beginners may also be missing the human engagement that helps hone that skill.

The quiet response seems poised to be far more discouraging than a few bits of negative feedback from the public. Silence gives you nothing to work with. You know you still need development, but you don’t know what or why.

So maybe they use AI to improve their writing. AI can help close the visible gap between taste and skill. The harder question is whether it sometimes closes that gap in the artifact faster than it closes it in the creator.

And even then, they may get flagged as slop. Not because the work is empty, but because our collective sensitivity to AI tells will keep evolving alongside the models themselves.

This isn’t about algorithms raising the bar and pushing out young artists. That may be true, but there may even be some benefit there.

When Everything Looks Finished

The real question may be what happens to creativity when the public spaces where we once learned in front of other people become increasingly optimized to show us only finished-looking work. What happens when we view every piece of mediocre developmental work as inhuman slop, not worthy of feedback aside from flagging it as “AI slop”?

I’m not saying we should stop looking for AI slop, or lower our standards and taste. That’s not how anything or anyone works, and it wouldn’t be helpful.

Maybe the problem isn’t that AI lets beginners make better-looking work sooner. That sounds mostly good to me.

Maybe the Slop Was the Lesson

The question is whether there were things happening inside all those embarrassing drafts that weren’t merely producing the draft.

Maybe struggling through the paragraph was teaching rhythm.

Maybe making the bad video was teaching pacing.

Maybe bombing in public was teaching what actually connects.

Maybe some of the slop was the lesson.

We should absolutely build systems that make low-effort content less rewarding. But we should be careful not to confuse low quality with low effort, or polish with value.

Some work is bad because nobody cared. Some work is bad because somebody hasn’t learned how to make it good yet.

Those people need somewhere to practice where the work can still be seen.

Ira Glass’s gap doesn’t disappear just because our feeds get better at hiding it.