AI is the wrong way to go looking for superintelligence. Far more realistic is developing means of organizing humans together effectively enough to achieve sup

AI is the wrong way to go looking for superintelligence. Far more realistic is developing means of organizing humans together effectively enough to achieve superintelligent levels of collaboration. I think before 2025 is quite reasonable given this approach.