if your customers die, they stop being customers. definitely some long term value there, but youre right, the psychological damage to the drivers could offset those gains.
Their replacement is automated transfer between van and package station.
Robot labor is far too valuable to waste on going door to door with boxes, humans will always be cheaper.
This is peak human narcissism. It won't be robots doing the jobs humans do not want to. It will be humans doing the jobs, that are so valueless that robots won't bother with them.
>Human Amazon drivers currently are around $40k per year to rent.
Now. Yes, NOW. Consider what will happen to the unskilled labor market when there are humanoid robots competing for these jobs. Going from place A to place B with a package is of so incredibly low economic value that it will always be cheaper for a human to do than for a robot.
Consider this. A robot at its lowest will cost around 10k with 2k to 5k yearly maintenance. Human labor at its lowest is around 3 dollars a day. Robot labor will never be cheaper than human labor.
In other words. A human consumes 2k kilo calories a day. How many calories does it take to produce, charge and maintain a robot? Without a doubt many orders of magnitude more.
To think that robots will ever be cheaper than humans is such a warped western perspective. You are just neglecting how cheap labor is around the world.
You can't globally arbitrage local labor, unless the labor is performed remotely, which the upcoming humanoid bots won't require much of in the delivery role.
Robot labor will first overtake where the greatest economic gain is. All human labor below that skill level will now be pushed down to lower levels.
Robot labor costs will be a hard upper limit for human wages for unskilled labor. There will be like 40% of the population in direct competition with robots. Human wages will drop hard and very fast. Thankfully things like package delivery are always cheaper for humans. Again minimum labor cost for humans is like 5 dollars a day. Robots can not compete with that, so the easiest unskilled jobs will remain human.
For these devices to be useful at all so many things would have to go right. The emergency must be reported, the van driver informed, he then has to drive (presumably without breaking any traffic laws, which EMS would do) to the affected person and correctly connect the device. The driver needs to be somewhat trained (although the devices are largely automated), but he also has to be prepared for some serious emotional distress. Having a person die right in front of you while you are fiddling with the medical device that could save his life is a terrible task. This is hard on EMS personell, but they are trained for this, have experience and self selected into that role. A van driver has none of that.
All in all this seems of dubious value. From the article: "Amazon confirmed that more than 100 contract drivers took part in the experiment, with several receiving alerts from citizen responder apps and arriving on site, where rescue services were already treating the victims." Meaning there was exactly zero value in the program.