Off Switch

You’re reading a post, an email, a doc — and then you hit it:

"It's not just a tool. It's a paradigm shift."

Your brain switches off. You felt it. Something about that sentence registered something as all-too-familiar. You stopped internalizing the idea and instead focused myopically on the construction. The writer lost you because you hit the intellectual off switch.

That construction has a name: negative parallelism. “Not X, but Y.” “Less about A, more about B.” “It isn’t the thing — it’s the other thing.” It shows up in a hundred variations, and every one of them triggers the same AI slop mental block: I want to connect with this idea but it smells like AI… and it should be simpler.

The data backs up your instinct

Will Oremus wrote the definitive piece on this for The Atlantic in July 2026 (link). The key stat: researchers at Pangram found negative parallelism appears three times more often in AI-generated text than in human writing. Barron’s reported its usage in corporate communications quadrupled between 2023 and 2025.

The model negates a weaker framing, then offers a stronger one. While the structure mimics insight, the reader’s semantic seesaw yields a yearning for simplicity: wouldn’t a simple direct sentence have been better? Why all of this negative parallelism tech bro AI slop? Why not just say that one affirmative, simple, direct sentence?

The prompt

Here’s the constraint I’ve started using:

## Language Constraints

- Do NOT use negative parallelism ("not X, but Y" / "less about X, more about Y" / "it isn't X — it's Y").
- Do NOT use antithetical phrasing that defines something primarily by what it is not.
- State what something IS. Describe it directly, affirmatively, succinctly, on its own terms.
- Avoid rhetorical seesaws that pit a straw version against the real point.

Isn’t that better? I’m calling this ‘positive synchronicity’, the inversion of negative parallelism.