Robot Sobriety: The Case of Amazon

The projected employment apocalypse at the company is greatly overstated

I laid out my case recently as to why the recent white-collar layoffs at Amazon were likely not about AI, and I’m in Jacobin today expressing skepticism about the idea that Amazon’s on the cusp of automating away its blue-collar warehousing and delivery workforce. Count me as an official AI skeptic!

The short of it is that I think the big automation displacement at Amazon is actually behind us. It happened during the pandemic, when Amazon’s package volume shot through the roof but its workforce remained fairly stable. They could do this because they aggressively pushed out mobile units (Kivas) during this time, which are the one clear robotics win for the company (the lion’s share of Amazon’s 1 million robots are Kivas). They’re hoping for some version of a robotics arm program to be the next wave of displacement, but I think there’s good reason to question the productivity gains there. And with regard to last-mile stuff, well, that’s just a silly show to titillate the investors.

Here’s the payoff of the analysis for union organizing:

To my mind, it’s important to be clear about this because there are any number of excuses I regularly hear about why Amazon is impossible to organize, one of the most prominent being that it’s on the verge of automating away huge swathes of its warehouse/delivery workforce. This is a thought that tickles investors and exercises a disciplining role on labor. Mercifully, it’s not true; unmercifully, it doesn’t need to be true for the company to continue on its path of roaring success, free of union penetration.

Full article here.