Relevant snippet from Palo Alto:

Safiyo Mohamed immigrated from Somalia only three months before she started as a stower at an Amazon warehouse in Minnesota. After three days of training (only in English), she began emptying incoming boxes from the conveyor belt into boxes from which the pickers picked. The system had a strict quota: 2,600 items sorted for every 10-hour shift, or a consistent average of less than 14 seconds per object. To fit in any kind of break you had to go faster, so as a beginner, Safiyo tried to avoid taking any breaks. Still, after her first week, her manager told her she was too slow and made more mistakes than the acceptable number: one per shift, an inhuman error rate of .04 percent. What help she got concerning strategies for improvement came from the other Somali workers who filled the warehouse. (As in the Hoover-era gold mines, white English-speaking managers boss groups of nonwhite workers, whose ethnic composition depends on the location; in Minnesota’s Twin Cities, a high proportion of them are East African immigrants.) “After my shift, I couldn’t even cook for myself. I barely had the energy to take a shower and often went to bed with an empty stomach,” Mohamed recalled in an essay about her experience for Sahan Journal. “I had nightmares about getting fired, disrupting the little sleep I was getting. They treated me and every other warehouse worker like a machine, not a human.” Feeling like she had no other way to support her family, Safiyo lasted longer than most, 30 months. In Amazon’s numbers, she turns up as an unqualified success. That’s how it’s supposed to work; that’s how Amazon came to dominate retail and how Jeff Bezos became the world’s richest man.

A detailed 2020 investigation found that Amazon’s warehouse workers have a serious-injury rate nearly twice the warehouse industry average. And the more robotized the distribution center, the higher the injury rate. “If you’ve got robots that are moving product faster and workers have to then lift or move those products faster, there’ll be increased injuries,” an Occupational Safety and Health Administration inspecting physician with experience in the company’s facilities told reporters. Racing to catch up with machines rigged to run hot hurts people. Efficiency causes injuries, which is another way of saying that, for Amazon, injuries are efficient. Amazon’s delivery must be very efficient, because drivers get injured even more—a lot more. And unlike warehouse workers, Amazon delivery drivers don’t technically work for Amazon, even if they’re wearing Amazon uniforms delivering Amazon packages in their Amazon vans following Amazon’s directions to Amazon customers.

For most of its existence, Amazon got its packages to customers’ doors just as everyone else does, contracting with the U.S. Postal Service or the big private shippers UPS and DHL for the dreaded “last mile” delivery. But as part of Operation Dragon Boat, the company planned to bring shipping inside the Amazon tent, or at least next to it. Driving trucks around is dangerous—for drivers, for the other people on the road, and for the legally liable employers. Trucks kill people, especially when you drive fast. Amazon drivers, unsurprisingly, have to drive fast if they want to keep their jobs, and their trucks do kill people. But when it comes time to hold someone responsible, Amazon is nowhere to be found. Patricia Callahan’s 2019 investigation for ProPublica and the New York Times couldn’t establish exactly how many deaths Amazon drivers were responsible for because they are all contractors, hired through the Delivery Service Partner program.

Amazon has absurd injury rates while they get a lot of great press for their “automation”, but what such “automation” is actually doing is setting a dangerous pace for the human bodies. With such high turnover and higher than local average wages, Amazon is able to pace workers at inhuman speeds, to their body’s breaking point, because each individual worker is an expiring commodity to them.