@skycolonthree@mia3as transformers, the core concept in all LLMs (Gemini, Claude, chat gpt which stands for generative pre trained transformer) have been introduced by Google in 2017 with the explicit task of language translation
@scheminglunatic@braearb but most maths used for LLMs and such is really really simple, like you can probably teach a middle schooler how transformers work in a week or less
@CatgirlAnarcho @azathecat chances are high they already have a good fingerprint on you through canvassing, basically rendering an invisible image and how it's rendered fingerprints your exact hardware and software combination
@badgraphix @Sadie11037 @HeresyL0ud It's the algorithm, one paper norms the latent space and early training test images give back piss yellow filtered noise images. That paper is referenced by open AI as well
@SHARKMODER@kiikazuu It's actually worse, he's saying that he hates losing his privilege and will vote against anyone trying to remove his privilege.
Actually crazy how he sees marginalized ppl being protected as a bad thing because it would reduce his privilege