Yeah, but how do you automate picking and packing of glass/crystal/etc, with the same equipment that picks/packs refrigerators? I mean AMZ deals in a ton of things of varying sizes; do you design different picking/packing equipment?
The answer seems obvious. One starts by creating a robot for a particular, limited job, perhaps to handle the most awkward job in the warehouse (like scuttling all the way to the opposite end for a particular item). Then either modify that robot for other jobs or create other robots for those jobs, or a bit of both. Eventually, logically speaking, one steadily needs less employees.
I'm not saying it would definitely be cost effective. IMO it's the sort of project that a company with a huge amount of financial resources at their disposal should be willing to pump money into on the chance that with enough development it will eventually be cost effective.
The economy has only improved in the state controlled media, not in reality. Entry level jobs are still close to impossible for kids to get, and there is nothing like a growth market even for experienced highly skilled people.
I'm not saying either of you are correct, but based on experience when a economical crisis hits, the low-end of the job market gets hit first. It would be logical if that's the first area that shows signs of recovery, IMO.