de Alexandru Giboi
Because one serious thing I was thinking about this past week was whether AI-hype is actually human-related. AI biases aside, reading all the problems that come from ungoverned or poorly governed AI products, I had a feeling that AI development contained a lot of projection of people’s frustrations.
What I mean is this: might be possible that LLMs like ChatGPT, while getting trained on all of the interactions with the users, also learn human behaviour patterns and also “ingest” our frustrations?
