Swimming in Slop

We’re drowning in AI-generated content, and the drop in quality is getting harder to ignore.

You can spot it immediately.

Hyper-confidence.
Generic structure.
Volume over insight. 

Em dashes as far as the eye can see.

What’s happening here? More importantly, what can we do about it, starting with our own contributions?

The tools aren’t to blame. We’re handing over the cognitive steering wheel.

And if I’m honest, I’ve watched myself doing it.

Here’s the progression I’ve noticed. The danger isn’t in avoiding AI or experimenting with it. The danger is in surrendering to it.

Stage 1:  Denial
We don’t use AI. Either we don’t trust it, or we don’t see the point. We do the thinking. That’s how we stay in control.

Stage 2:  Reluctance
We start experimenting. A tricky email. A quick summary. A better vacation plan. It’s useful. Occasionally impressive. But we’re still in charge. AI is a tool, not a participant.

Stage 3:  Deference
This is the inflection point. This is where the slop begins. AI feels like a superpower. Speed goes up. Output goes up. Confidence goes up. But something else quietly drops: the quality of thinking. We stop interrogating the output. We start accepting it. We confuse fluency for insight and polish for judgment. This is where the slop comes from. Not bad models. Unexamined adoption.

Stage 4:  Challenging
This is where you take the steering wheel back.  A more deliberate stance emerges. You don’t reject AI. You don’t defer to it. You engage it. You assume the first answer is incomplete. You probe it, pressure-test it, and rewrite it. You use it carefully. You don’t let it replace your thinking. You say “no” as often as you say “yes,” and you know why.

The process may be faster, but it’s not easier. The work has shifted from producing an answer to taking responsibility for it.

Stage 5:  Harmonizing
You stop chasing better answers and start designing better thinking. You decide what gets delegated and what doesn’t. Where AI accelerates you, and where it has no vote. You don’t just respond to the output. You structure the exchange.

AI becomes useful not because it’s smarter, but because you are more deliberate.

The goal isn’t to stay in control. It’s to lose control, notice it, and take it back differently.