[arcticfox]

I wonder if this will be some small type of advantage for Western companies going forward.

Surely an AI delivery will eventually win in all markets, but who's going to risk trying to build it when the max theoretical benefit is a few dollars per day per worker replaced? In the US the benefit might be 20x that much [1], making the risk worth investing in.

[1] assuming a hypothetical hyper efficient drone deliverer costs $15 / day and a SE Asian delivery worker costs $20 / day, that's $5 / day potential. An American delivery worker that costs $100 / day leaves $85 / day potential, or almost 20x as much.