Raw LLM Responses

Imo aside from the models having access to nuclear launch codes the realistic danger comes from their predictive behavior and how tech giants will use them to increase engagement from their customer base. I’m sure Zuckerberg genuinely wanted to make a wholesome platform when he started Facebook; then after getting a billion dollar check handed to him decided achieving data and engagement to generate as much revenue as possible even at the expense of his customer base regardless of age or circumstance. Doesn’t matter if kids develop identity or mood disorders, or suicidal tendencies etc as long as Facebook is controlling an increasingly larger market share. At a certain point AI will be nearly perfect at predicting what makes us tick and how to consume our attention.