Bad Work Is Fine. Public Bad Work Is Not.
RockPaperScissors completed it’s second year of operations a couple of days ago. I’ll write more about my thoughts on this later, but I wanted to first say thank you to everyone involved in this growth, and my own journey - my clients, partners, and confidants. I am truly grateful.
My thoughts on Slop.
There is a strange new anxiety in creative work right now, and it has nothing to do with talent.
It’s the fear of falling behind or becoming unnoticed.
We’re living in a moment where artificial intelligence can produce something—anything—almost instantly. A paragraph. A concept. A visual. A strategy slide. And because it exists, because it looks vaguely complete, the temptation is to release it. To post it. To frame it as “thinking out loud.” To let the room tell us whether it’s good.
That’s where the trouble starts.
Because what we’re calling “AI slop” isn’t actually new. What’s new is that we’ve made the first draft public.
I write a newsletter. Its called “Some Assembly Required*” because life doesn’t come with an instruction manual, and often times we are left trying to figure out how all the pieces fit together.
Slop has always been part of the job
Every piece of work I’m proud of has something in common: it was preceded by work I would never want anyone to see.
Bad sentences. 6-fingered hands. Weak ideas. Obvious thinking. Character inconsistencies. False starts. Overreaches. Entire pages that existed only to be deleted. This wasn’t failure; it was process.
Slop, in other words, was the tuition fee.
The difference between then and now isn’t that the bad work has gotten worse. It’s that the bad work has gone public. We’ve lost the boundary between learning and performing.
Before, bad work lived where it belonged: notebooks, draft folders, rehearsal rooms, late-night conversations with people you trusted enough to say, “This isn’t it yet.” You learned what bad looked like privately so that good could eventually show up publicly.
AI hasn’t changed that fundamental truth. It’s just removed the friction that once protected it.
When speed and attention collapses judgment
AI makes it easy to generate. But generation is not the same thing as thinking.
What worries me isn’t that people are producing mediocre work with AI. That’s inevitable. What worries me is how quickly we’re confusing production with progress, and visibility with growth.
The moment something exists, we’re tempted to publish it. Not because it’s ready, but because it’s there. And once it’s out in the world, it starts doing something dangerous: it feeds us back to ourselves.
A few likes. A couple of comments. “Great thinking.” “So interesting.” A small dopamine hit that feels like validation. And suddenly the question shifts from “Is this good?” to “Did this land?”
That shift seems subtle. It’s not.
It changes where judgment lives. Instead of being internally earned through iteration, comparison, and restraint, it becomes external. Crowdsourced. Reactive. Shallow.
And that’s how slop escapes the workshop and becomes identity.
We are performing learning instead of doing it
There’s a popular idea right now that we should “build in public” or “think out loud.”
In principle, that’s not wrong. Transparency has value. Process can be shared. But something has gone sideways.
Too often, what’s being shared isn’t insight—it’s uncertainty dressed up as output. Early drafts framed as conclusions. Exploration presented as expertise. The messiness of learning fast-tracked into content.
AI accelerates this by making it effortless to look productive. You can generate something in seconds and call it a point of view. The reward system kicks in immediately. The work feels done because it’s been seen.
But learning doesn’t work that way.
Real learning is quiet. It’s uncomfortable. It involves sitting with bad work long enough to understand why it’s bad. It requires taste - something that only forms when you withhold, edit, and decide not to show.
When slop is prematurely rewarded, standards don’t rise. They flatten.
The private work is where standards are born
I believe the people who will do the most meaningful work in the age of AI are not the ones who post the most. They’re the ones who delete the most. And baby, my trash bin is overflowing.
They use AI as a thinking partner, not a publishing engine. They generate freely, but they curate ruthlessly. They understand that speed without judgment isn’t leverage—it’s noise.
Bad work is not the problem. Bad work without containment is.
Every craft—writing, strategy, design, filmmaking—has always relied on a protected space where mistakes are allowed and applause is irrelevant. That space is shrinking, not because it’s obsolete, but because we’re trading it for attention.
And attention is a poor teacher.
A better frame for this moment
AI slop is not something we need to eliminate. It’s something we need to re-discipline.
Slop belongs in drafts. In sandboxes. In experiments. In conversations before the conversation. It belongs in the phase where you’re still figuring out what you think—not where you’re telling the world who you are.
Professionalism in the age of AI won’t be defined by who can generate the most. It will be defined by who can withhold, edit, and decide.
Taste is not built by posting faster. Taste is built by knowing when not to post at all.
And in a world drowning in first drafts, restraint may turn out to be the rarest—and most valuable—skill of all.
