a lot of applications for humanoid robots i've seen people make other accurate and , despite being very complex, much more simple machines for.
they are for a specific purpose ofcourse, where such a humanoid robot might be more general purpose. but general purpose things usually mean less good at specific things.
if it becomes good at everything and cheap enough it invites the next more economic problem (which i think is more real than 'AI' simply doing it. its robotics that will do it..) that people will lose jobs. those people cant purchase things anymore which their replacements produce, and thus there is a problem that now the robots do nothing anyone needs anymore?
i can think of specific jobs no one should need to do, like handling nafta or other insane scenarios, radioactive stuff etc. but those are so specific you could think a specific robot would be better at it and theres no need for a general purpose device to do that specific job.?
seeing how much is invested in this, i am quite sure i am missing an insight, and would be thankful if someone in robotics applications might enlighten me!