Personally, it sounds pretty dystopian to me to live in a world where all labor essential for human living could be automated but isn't, stagnant because inefficiencies in society must be forced to prevail above everything.
Richard and others fail to see how a *general labor machine* is qualitatively distinct from every other form of automation that has been developed in history.
We have historically automated *tasks*, but not *labor* itself
The question of whether AI can actually become the Universal Labor Replacing Machine remains an open oneβitβs certainly not there yetβbut the implications for a world in which it *did* are very obviously different than with previous automation technologies