Daniel AppelquistIf AI is so great, why does it always seem to be on the side of fascism? https://www.bbc.co.uk/news/articles/c5yj97yd6v5o
Marijn@torgo Don't ascribe it agency. It's not on any side, it's spitting out what it's asked to spit out. And the fascists are, unsurprisingly, the group most enthusiastic about the idea of automated lie machines.
MycotropicI'd say also that fascism is baked into LLMs not by design or bad faith programming, rather by the basic nature of a neural network/LLM. It's designed to summarize, simplify and return generalizations based on weights. Semantic ablation is a feature not a bug and fascism is an intrinsically simplistic way to classify; us/them, good/bad, right/wrong it fits in perfectly with simplifying and summarizing.