You might think that grocery stores in poor neighborhoods have lower prices than wealthier neighborhoods (because poorer customers are more likely to shop around for bargains). The opposite is true. Wealthy customers have the time, energy and money to be able to shop around. Poor people are stuck in "food deserts".

Some thought experiments for you:

1. This store is located between your employer and your home. The time is "right after typical working hours". The presumption that the algorithm can predict is that you're picking something up from the store to make for dinner, therefore you have no time to price shop and therefore we can charge you more because you have to have it now.

2. American shoppers tend to have very high brand loyalty. If you always buy product_A and never product_B, then the algorithm can charge you more for product_A because your demand is quite inflexible.

This post of mine from a few days ago has links you might be interested in: https://news.ycombinator.com/item?id=47848128

> I'm genuinely shocked that ill-defined "predatory pricing" is the issue we see in the news.

There are lots of evil people out there working very hard at extracting more than their fair share of wealth from others. Far too many people believe in the "Just World Hypothesis" - that "what goes around comes around". There are lots of random bad things and bad people that hurt innocent people. There might actually be some positive benefit that could come from surveillance pricing. However, once you've known or worked with some of the more predatory people then you'll see just how hard this technology can be used to screw others.