Earlier quoted context omitted.
There is a finite number of things to do which create value and which humans are willing to pay for (just as is the case with horses). The only difference is that N_human > N_horse. The fact that N_machine < N_human doesn't mean that N_machine will always be less than N_human. I'm only pointing out that it is a fallacy to claim that it is.
In human society, there are a practically unbounded number of things that humans can do that other humans would be willing to pay for. The same cannot be said for horses, which have zero ability in human society to find new ways to create value for humans and thereby for themselves. To use your notation, when N_horse slumps toward zero, no horse can do anything about it. But when N_human declines, the displaced human…
Could you explain this statement further? Why do you believe the number of things humans can do is best modeled as being infinite?