There is a carousel page on the V3R website(The Print Conversation). The first line of it, until very recently, read like this: "What you can feel, you remember. What you remember, you buy." Underneath, in attribution: "The whole game, in eight words."
The quote isn't eight words. It's eleven.
I caught it before posting it to LinkedIn. I screenshotted the slide, sent it to the AI that had written it, and asked: spot anything wrong with this?
The AI's first answer was that the Previous button had a stray gold border, and was that the problem?
It wasn't. I told it no, look again.
Second answer: the quote was nine words, and the attribution was wrong.
Nine. Still wrong. I'd asked an AI to find a counting error in a quote about counting words, and it had managed to miscount the words while doing so. Twice.
The third time, the AI counted them out one at a time and arrived at eleven. We rewrote the attribution. The carousel now reads "a brand client, after seeing the foiled cover."
What this series actually is
I want to be honest about how the seven essays that follow were made. Every one of them was written by an AI. Not "polished by AI." Not "AI-assisted." Written.
What I supplied was thirty years of opinions about what the print industry doesn't talk about. A year of muttered mantras to anyone who'd listen — about visualisation, about the gap between what designers can specify and what brands can picture, about how embellishment work dies in the brief stage because nobody can show it. Three Color-Logic research PDFs full of market data I knew was true but hadn't sat down to write up.
The AI turned all of that into essays.
And along the way, I caught it being wrong. Not occasionally. Constantly.