@ESYudkowsky 🤔 more context, I think I'd been listening to Liron in a Spaces and commented in the GC that idk why he's insisting on arguing *literally any* goal when even a weaker form that just says many goals are compatible w superintelligence is enough to make the point they're not all aligned by default... I'm probably not the source of this, but if I am, I didn't intend to characterize your arguments, I was reacting to Liron & others and using strong / weak as the first words that came to mind to distinguish "most" from "literally all"