The Machine Doesn’t Get a Byline

We use these models, and we are not going to sell them to you.

Somebody who wants help with a manuscript wants their repetitions found and their commas fixed. “AI-powered” names the machinery and asks you to be impressed by it; proofread says what you get. So we name the work and never the mechanism—which is not concealment. The documentation says plainly what runs and where. It is a refusal to sell you the machinery as though it were the benefit.

The badge on the other side is the same move in a different coat. “Made entirely by humans” also trades on how you feel about machines rather than on whether the thing is any good, and we will not wear that one either. Nor will we tell you what to think about any of it, or make you feel foolish for using what you use.

There is one line the tools don’t cross, and it is about what comes out rather than what goes in. A model reading your draft back so you can hear what your eye skipped is a tool: one listener, and the listener wrote it. A model reading your book aloud for strangers is a performance, and a performance is a job somebody does. The test is whether the thing it made leaves the room.

What crosses is answered upstream and does not need restating here: nobody you did not name gets a copy, and nothing you make is training data, ours or anyone’s.

And the cost, since the fashion runs the other way. We cannot promise all of this happens on your computer. On-device inference is a promise only new hardware can keep, and we build for the laptop you already own—so some of it will run on ours, named, disclosed, and holding nothing once it is done.

The machine helped. It doesn’t sign.