I feel like machine learned models of language and sentiment are all knee-jerk intuition with none of the principled introspection


that is to say, your AI has internalized transphobia and you don't have the tools to make it unlearn it

when youtube autoflags a video as violating community guidelines, and removing the word "trans" from the title fixes it, I'm certain there's an LSTM somewhere in the chain that ingests metadata and basically says "cringe" or "based"

people do the same thing all the time. that's cringe culture. it's comfortable, trained disgust. no moral thought needed. the only difference is computers can do it now


Someday, historians will use our machine learning models as a way to understand the zeitgeist of our era.

Sign in to participate in the conversation

cybrespace: the social hub of the information superhighway jack in to the mastodon fediverse today and surf the dataflow through our cybrepunk, slightly glitchy web portal support us on patreon or liberapay!