[nullfield]

There seem, to me, really two bifurcating paths ahead. Maybe 2.5.

In both of them, AI is Not My Problem.

In A, we get a technological singularity, and AI is beneficial and benevolent; life is radically improved. In A.5, we don’t quite get to that technological singularity immediately, but get vast improvements in domain-specific systems from individually optimized models. Life is still significantly improved, so much that the edges of post-scarcity start to hit. This is the only one with a maybe caveat, but if anything it’d be leveling disparities.

In B, we get a technological singularity (or closeish) but fail to get the alignment problem down. The Terminator movies are so far from how badly we’d lose it won’t matter.