"Controlling models may not be the right answer. The two lenses of belief and evidence lead us down different paths...while responsible deployment will help prevent incidents, alignment may not prevent humans from wreaking havoc with even unintelligent AI."
blog.dfeldman.co/navigating-t...
blog.dfeldman.co
Navigating the AI Insanity
AI is exhausting—or at least the discourse around it is exhausting. It seems to get hypier and weirder by the week. I wish I could ignore…