Another thought.
Human bodies are a good jack of all trades master of none.
The most efficient, particularly, factory robots don't look human.
Amazon's robots that move around a flat floor and move racks of goods is one example. They can't pack boxes or label them, but they are the beast at moving racks.
A truly efficient factory will be designed around tasks and it's robots will too.
They won't be designed around humans.
You don't have to be human to pack a box, for example.
Biscuit factories don't have rows of humanoid robots with cookie cutters, they have a dedicated machine to cut them out.
Humanoid robots in factories have little use, except in low volume or frequently changing, or trial environments.
It will be a fad for the most part.
Factory owners/designers will quickly change them to dedicated robots for each task, where there is enough frequent use.
Why use a $20,000 USD humanoid to do a simple task a dedicated robot can do for far less money.
For example a robot packing a box at a stationary conveyer doesn't need legs. Yet a humanoid has them. So it's more to go wrong with, wear out, malfunction, extra initial cost, so on.
PS
Yes, Amazon has a humanoid robot, like everyone else.
The owner also mentioned losing billions on ideas that haven't worked out.
They have it moving stuff from one side of a small room to the other.
A robotic arm hanging from the ceiling or mounted on the floor would be cheaper and more efficient.
Could probably move faster too.
I think it's more marketing then anything else.
PPS
I see a spike in crime as many low end jobs are suddenly gone.
Think fruit picking, cleaning, retail.
Much cheaper to replace staff with a robot. They will pay for themselves within a year or so.
Again the staff won't just sit around.
If not crime, they will just start human businesses that will likely beat the robot competition, once the novelty wears off for consumers in about two weeks.
