Also, no one is talking about how internet content providers are going to intentionally switch from writing for human audience to writing for LLMs. The articles on the internet will still be their, bur their main purpose will be to provide training data for LLMs in order to get the LLMs to push their message. Want an LLM to mention or suggest your brand when a user asks about a specific product or problem? Mass-create large quantities of LLM readable website text so that when the LLMs scrape the web for training data, they get trained to promote your product.
Meanwhile, the paranoid creators of human-centered content that may have more impartial and truthful data get paranoid and do their best to HIDE their content from LLMs behind "prove you are not a bot" cloudflare walls, because, "Oez noez!!! LLM might get trained on my article amd that'd be terrible, because how dare it!! Stop it immediately!!!"
Yeah... The internet is about to go through a phase.... 😒
buzzbot (nprofile…k5h0) 10000
A look into how X’s Grok AI started replying to everybody about white genocide.
After this blog was published, xAI, who own X, started trying to hide the evidence on GitHub.
https://smol.news/p/the-utter-flimsiness-of-xais-processes
