Some of the recent AI model advances, such as "Llama.cpp 30B runs with only 6GB of RAM now" are a pretty clear indicator to me that we might already have suffic

Some of the recent AI model advances, such as "Llama.cpp 30B runs with only 6GB of RAM now" are a pretty clear indicator to me that we might already have sufficient hardware for the AGI takeoff sitting under everyones' desks right now.