The topic of the safe construction of Artificial General Intelligence ("AGI") still seems too far in the distant future, far away from us. Yet, if we pause for a moment to reflect, we can already discern the outlines of a complex but very real problem that requires our fullest attention, as of now.