The concept of “deep content” is not as important as the commonality of the topic when it comes to protecting content from AI-powered content generators.
These generators, which are mostly based on large language models (LLMs), work in a statistical manner.
Rare topics are generally safer from being replicated by AI.
One reason to include “Experience” in content, as part of the EEAT framework, is to make it more unique and less common compared to the vast corpus of internet documents that AI models are trained on.
However, it’s not just the scarcity of a topic that matters, but also the scarcity of “good” content on that topic.
The more low-quality, weak content there is about a particular subject, the better a high-quality piece on that subject will be hidden from LLMs, especially those that don’t search or scan and instead take the top results from search engines.
When evaluating content, it’s important to consider not only factors such as format, search volume, business value, topic breadth and depth, and content value but also the quantity and quality of competing pieces.
The more fluff and low-quality content there is on a topic, the less likely AI is to steal original, high-quality content on that topic.