These are deterministic, source-backed cards from the offline export. They are not live AI answers.
LLM outputs can be influenced by optimized assets that appear in web search results, including press releases, Reddit, YouTube descriptions, and social posts.
@@build_in_public · asserts
I just want to remind you it is incredibly easy to influence large language models. People are using press releases, Reddit, YouTube to rank number one on Google in hours. Large language models use web searches to get most of their information. They come across what is being planted by SEOs. They regurgitate it as if it is fact. They don't actually know....
I just want to remind you it is incredibly easy to influence large language models. People ar...
@@build_in_public · 2026-05-20
I just want to remind you it is incredibly easy to influence large language models. People are using press releases, Reddit, YouTube to rank number one on Google in hours. Large language models use web searches to get most of their information. They come across what is being planted by SEOs. They regurgitate it as if it is fact. They don't actually know....
I just want to remind you it is incredibly easy to influence large language models. People are using press releases, Reddit, YouTube to rank number one on Google in hours. Large language models use web searches to get most of their information. They come across what is being planted by SEOs. They regurgitate it as if it is fact. They don't actually know. They just say what comes up. This is a specially useful for reputation management. Somebody's doing due diligence on a brand. The LLM sees the description of a...
Cookie preferences
We use necessary cookies to run the site and optional cookies to understand what pages are useful. You can accept all, reject non-essential cookies, or manage preferences.