The biggest problem with robots for a long time has been the brain, but now with LLMs, and other things, being good enough to fully control robots, even if using multiple specialized AIs, the tendency is that more and more will be invested in this direction.
Yes the first generations of robots might be slow and have problems, but that will be used to train the newer AIs for both embodied tasks and also for fully in-computer based tasks, meaning that by the time they can finally be produced cheaply and in mass they will be ready for many jobs.
This may also soon lead to less humanoid designs, like robots with multiple arms and no legs, leading to both generalist and specialized robots that can work very fast to take all manual labor jobs with ease in decades or less. Hopefully the people in the capitalist countries can manage to organize to work for changing the system into one where the people can benefit from the full automation that is to come.
The new average of global robot density in the manufacturing industry surged to 141 robots per 10,000 employees – more than double the number six years ago.
The following is a very basic calculation with many assumptions, of course, but if it continues doubling in six years as stated that would be less than 40 years for full automation. And that's before taking into account more focus on making robots as they get better or robots getting better to the point they can make more robots faster and cheaper by themselves.
