This means that in the not so distant future robots will no longer require annoying programming. Using machine vision, motion sensors, image recognition and voice recognition, and advanced new software, they will be able to cope with ever more intelligent tasks, including communicating with the environment and constantly learning from it, and especially from people .

The only issue here is that all of this requires programming. Machine vision (CV) is also a big pain in the ass, but it can do a lot of good stuff if you have the right team behind the code.

I think I get what you're saying though: less on the spot coding will need to be done. That's correct, IMO.