See my obviated reply to cousin_it: yes, the physicists are wrong to make this inference, and it's because they are not using the appropriate model-to-reality mapping, which makes its usefulness appear questionable, and this negative appearance is further worsened when they get to the point of literally saying "Humans holding heavy stuff isn't work." (Which is wrong in both the lay and the technical sense.)
It seems to me that usually, when someone says "ethics" on lesswrong, ey usually means something along the lines of decision theory. When an average person says "ethics", ey is usually referring to a system of intuitions and social pressures designed to influence the behavior of members of a group. I think that a lot of the disagreement regarding ethics (i.e. consequentialism vs deontology) is rooted in a failure to properly distinguish between decision theory and what society pressures people to do. Most lesswrong users probably understand the distinction fairly clearly, but we only ever talk about decision theory. Why don't we talk about the social meaning of ethics?