43. Slop Didn't Start with AI
It has always existed
We talk a lot about AI shipping slop. Take a look at any enterprise software, and you’ll see the slop humans have shipped over the years, slowly and inefficiently.
Most of it comes from people with no skin in the game. People who don’t care about outcomes because they’re not measured on outcomes. AI will get better. Mindset is harder to change.
AI just makes the slop more visible. It accelerates whatever incentives already exist. In the hands of teams that care, it removes busywork and sharpens focus. In the hands of teams that don’t, it produces faster, shinier slop.
Good software has always come from ownership, taste, and accountability. That doesn’t change. If anything, the bar gets higher. When building becomes cheaper, judgment becomes the constraint.
This Isn’t New
Shipping mediocre software isn’t new. Slop didn’t arrive with AI. It’s been around forever. Not just in software either.
Look at most Netflix movies. Endless content, forgettable plots, safe characters, zero risk. A large chunk of those scripts could already be replaced by an LLM and no one would notice. That’s not a statement about how good AI is. It’s a statement about how low the bar already was.
Slop Is an Incentive Problem
If a team ships slop with AI, they would have shipped slop anyway. AI just removes the “it was hard” excuse.
I’ve watched big teams spend months building features that got discarded. Useless work dressed up with meetings, status updates and slide decks. This is what happens when you incentivise optics over outcomes. When “looking busy” matters more than “being useful.”
The things that provide optics, like meetings, processes and alignment sessions, don’t necessarily lead to good outcomes. They just look like work. AI helps you get to bad outcomes faster. It exposes the lack of skill that the process used to hide.
The case for small teams
Yes, AI floods the world with average output.
But it also lets small teams with taste compete against large teams without it. When shipping becomes cheap, taste becomes rare. And taste and care don’t scale through headcount.
A two-person team that knows what they’re building can now outship a fifty-person team that doesn’t. The bottleneck isn’t execution anymore. It’s knowing what’s worth executing and caring about the outcomes and customers.
Slop is not the same as shipping fast
There’s a difference between shipping something imperfect because you don’t care, and shipping something imperfect because you want to learn. The first is slop. The second is craft.
Slop is careless speed. They look similar on the surface, but the intent is completely different. One ships to get it over with. The other ships to start a conversation.
AI makes this gap more obvious. It lets you run the learning loop faster. Ship, listen, adjust, ship again. But if you’re not listening, you’re just making noise.
The one-day test
Here’s a simple filter: What’s the most valuable, most imperfect version of this we could ship in one day?
Slop-shippers can’t answer that question. They don’t know what’s valuable. They just know what looks like work. They’ll spend weeks on something that didn’t need to exist.
Teams that care can answer it immediately. They know what matters. They know what to cut. The one-day constraint forces clarity. It separates the people who understand the problem from the people who are just filling time.
Fewer excuses
For years, you could hide behind “it takes time” or “we need more resources” or “the technology isn’t there yet.” AI takes those excuses away.
What’s left is just the work. And whether you actually cared about making it good.
Start shipping and care about what you ship.