Join Nostr
2026-06-16 10:41:39 UTC

Pepijn on Nostr: I too love dunking on LLM tools. With this type of claims I do wonder whether this is ...



I too love dunking on LLM tools. With this type of claims I do wonder whether this is really new behaviour, or simply a new shape of something old.

In 2005 I created a Wikipedia page about a topic edging environmental protection and law. It was factual but did include a key bit of editorialised wishful thinking.

Soon after there were many language versions of the page, and a multitude of NGO and government websites boosting and acting on the claim.. in reality the law was only enacted in 2020.
New: Researchers have quantified how easy AI search is to manipulate. Just 13 words buried in a random Reddit comment can poison AI search results. They suggest this is not easy to stop: "The way you can attack these systems is so much dumber than you think it is"

https://www.404media.co/it-is-trivially-easy-to-use-reddit-to-manipulate-ai-search-research-suggests/