I’m not sure I understand your point. As far as I can tell, your objection is that labor productivity is mostly a function of level of automation, and you’re t

I’m not sure I understand your point. As far as I can tell, your objection is that labor productivity is mostly a function of level of automation, and you’re thinking people are confused and think it means “the people worked harder/smarter but weren’t allowed to make new machines to achieve that goal” or something like that? In the specific example you give, I’m really unclear why it would be bad to transition people from a mind-numbingly-boring-but-requires-constant-attention job working on an assembly line, to an essentially social job with a modicum of politeness/effort required. I guess I agree that this is fundamentally one of the flaws in Graeber’s analysis; if people’s only labor force options are “self actualized helping people with lot’s of autonomy and reasonably good pay” vs. “forced to serve others for essentially feudal reasons with slightly better pay”, obviously many if not most people would choose the former. But the reality is that the alternative to a service job are often assembly-line boring. It’s not obvious to me how many people would choose assembly-line over feudal retainer. Of course, there’s an alternate objection to analysis based on labor productivity, but it’s fundamentally Marxist: complete automation gives capital vastly more leverage and labor nearly zero. So yes, labor productivity says nothing about distribution of the gains from automation and is mostly about automation, not meaningfulness/autonomy for the worker.