> ## Content Index
> Fetch the complete content index at: https://globalfeed.ai/llms.txt
> Use this file to discover other available public pages before exploring further.

# Expert intelligence as the antidote to AI slop: Pinker's book reviews your draft
- URL: https://globalfeed.ai/en/expert-intelligence-as-the-antidote-to-ai-slop-pinkers-book-reviews-your-draft/
- Published: 2026-09-01T13:42:37.000Z
- Updated: 2026-09-01T13:42:37.000Z
- Description: Steven Johnson, editorial director of Gemini Notebook, showed the pattern: add Steven Pinker's book on writing to a notebook and have your draft critiqued through Pinker's principles, with citations.
- Author: GlobalFeed Editor
- Tags: google, gemini notebook, ai, x-stevenbjohnson, x-Gemini_Notebook, x-sapinker, dil-en, elle, video

An interesting answer to the flood of mediocre AI-generated text came from inside Google. Steven Johnson, editorial director of Gemini Notebook and Google Labs and the author of 14 books, demonstrated what he calls expert intelligence on his own draft: he adds Steven Pinker's modern writing classic The Sense of Style to a notebook, then asks for feedback on his draft based on Pinker's principles.

In Johnson's words, it is the opposite of AI slop: personalized advice grounded in human expertise, merely facilitated by AI. In the example he shared, the system criticizes a film digression that swallows a third of his essay by invoking Pinker's principle of proportionality, backing the diagnosis, the reasoning and the remedy with citations to specific passages in both Pinker's book and Johnson's own draft. The bonus he highlights: the suggestions are actionable, and the writer absorbs Pinker's underlying philosophy of writing along the way.

There is a product side too: a featured notebook curated by Pinker is available to everyone, and those who buy the book can add it to any notebook and run their drafts through the expert of their choosing. The formula generalizes: whatever field you write in, put its classic into your notebook and have your draft read through that master's eyes.

Johnson is no ordinary user; he has over a million followers, is followed by the likes of MIT CSAIL and Simon Willison, and leads the team that owns this feature. Still, the pattern he showed ties into the context debate we covered this week: getting good results from AI is about feeding it the right material, not more material.