Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The game theory of how algorithms can drive up prices

quantamagazine.org

Post reply on HN
Print view View original

111–120 of 146 posts

Re: The game theory of how algorithms can drive up prices

#111

Post by doctorpangloss »

RataNova wrote: post #95
pixl97 wrote: post #5

I mean we are in the age of digital pricing, even on the shelf. Modern price collusion is more apt to happen with A/B testing if prices at locations to see what the local market will bear. I've seen Walmart do this in the past. Items that were not on sale could have significant differences in price, where in general the prices in more affluent areas are higher. We're talking 50 to 75 cents on common items, but sporti…

Yeah, dynamic pricing isn't just for airlines anymore, it's quietly baked into everyday retail, and most don't even notice

I don't know. Surely the biggest impact on Amazon's prices, for example, is that it is disguising a $7 shipping cost inside the $0.10 items that are priced for $8? Or disguising $100 of prepaid shipping as a "Prime" membership?

Re: The game theory of how algorithms can drive up prices

#112

Post by hn-reader-2 »

wat10000 wrote: post #54
spwa4 wrote: post #44

Earlier quoted context omitted.

And yet, if you check it out for real, you'll find most food could be a lot cheaper (some countries have regulations for basic foods to be excepted from most regulation and taxes, and there's a large price difference) Especially meat could be a great deal cheaper if these countries wanted to make that happen. Food in the west is only cheap in one sense of the word, and even then if you compare how much of the cheapes…

Agricultural productive capacity hasn't changed that much in the past 25 years. Looking at the longer term, food prices have dropped enormously. At the beginning of the 20th century, the average American household budget was 40+% food. Today it's around 10%.

>the average American household budget was 40+% food. Today it's around 10%.

Does that mean food prices have dropped enormously or could it be that families have to spend more money on eg. rent, gas, and health? Adjusting for inflation, the price of milk have only decreased 1.1%[1]

[1] https://www.usinflationcalculator.com/inflation/milk-prices-...

Re: The game theory of how algorithms can drive up prices

#113

Post by mlsu »

cogman10 wrote: post #105
mlsu wrote: post #100

Earlier quoted context omitted.

This is it. and: true across the board, not just restaurants...

Yup, and I really hate it. Monopolies and oligopolies are really terrible in just about every way imaginable. Everyone that isn't an oligopoly gets screwed. This is also simply the natural end state of free market capitalism. Every one of these giant businesses knows that by swallowing up smaller competitors they can ultimately improve their revenue without improving quality or actually innovating/competing. Companie…

it's just incredible.

Internally, these huge corporations behave exactly the same as a good old fashioned USSR bureaucracy:

endless meetings where no work gets done

a huge class of bureaucrats (manager, senior manager, VP, senior VP, director, senior director ... what's next? commissar? secretary?) who don't actually do any of the line work and instead exist only to perpetuate a process

huge amount of process that does nothing for the bottom line or indeed for anyone at all

random party lines that you must accept or be fired (new director came in. Now we're doing a 30 minute velocity retrospective every week. you must attend, comrade!)

party language determined from on high, that changes once every 5 years (blockchain is our five year plan! huh? blockchain? no no, AI is our five year plan!)

party princelings who rise not on merit but purely on positional signifiers alone (Comrade, I know you've been a party loyalist for 25 years, but your senior director position is being given to a new princeling. He's 26 years old. He came from Stanford, and was on the forbes 30 under 30. They say he was a protege of Peter Thiel!)

and, most importantly: everyone at the bottom, who pays for all of it, and must take it completely seriously.

Re: The game theory of how algorithms can drive up prices

#114

Post by pessimizer »

bfkwlfkjf wrote: post #21
nomilk wrote: post #6

The researcher says > this strange strategy will maximize your profit. “To me, it was a complete surprise” It doesn't seem like such a surprise that algorithms that use information about rivals to optimising profit tend to price high. Consider a small town with two gas stations, you own one. You can set the price (high or low) in the morning and can't change it until the next day. Your goal is to optimise profit for…

> what can regulators do? Regulators could say "you're not allowed to make more than X profit". They already do that with utilities, so it's not a matter of practical impossibility.

I continue to believe that in the case of oversized margins, the government should just enter the market themselves. Buy the smallest competitor and operate it at a reasonable margin, growing it at every opportunity. If the rest of the market lowers their margins to beat it, spin the thing off.

Basically don't bother to dictate margins, just declare that market a failure.

Re: The game theory of how algorithms can drive up prices

#115

Post by cogman10 »

mlsu wrote: post #113
cogman10 wrote: post #105

Earlier quoted context omitted.

Yup, and I really hate it. Monopolies and oligopolies are really terrible in just about every way imaginable. Everyone that isn't an oligopoly gets screwed. This is also simply the natural end state of free market capitalism. Every one of these giant businesses knows that by swallowing up smaller competitors they can ultimately improve their revenue without improving quality or actually innovating/competing. Companie…

it's just incredible. Internally, these huge corporations behave exactly the same as a good old fashioned USSR bureaucracy: endless meetings where no work gets done a huge class of bureaucrats (manager, senior manager, VP, senior VP, director, senior director ... what's next? commissar? secretary?) who don't actually do any of the line work and instead exist only to perpetuate a process huge amount of process that do…

It's always wild to me when people talk about the efficiency of private industry.

There's not a private business on the planet that's super-efficient.

Re: The game theory of how algorithms can drive up prices

#116

Post by Workaccount2 »

Arwill wrote: post #91
shmatt wrote: post #71

Earlier quoted context omitted.

This actually works well the other way around. When sales are still growing YoY (like the post covid market), but prices are up 30% or 40%, you understand your customer is still willing to pay the higher price Its similar to a McDonalds or Starbucks situation where you just keep increasing prices dramatically until you get a first quarter of lower than expected sales, then you start adapting downwards Most corporatio…

>see streaming companies increasing prices every few months They can do that because they are practically monopolies.

They can do it because people are hopelessly addicted to screens.

You won't die if you stop watching Netflix. We aren't talking food or medicine here. In fact your life would probably improve. But addiction is a real animal.

Re: The game theory of how algorithms can drive up prices

#117

Post by Workaccount2 »

BizarroLand wrote: post #67
positron26 wrote: post #56

Earlier quoted context omitted.

This is still so oversimplified. There's always bellwether products customers buy a lot and get used to. They use those to decide if you are cheap or expensive. Costco hotdogs are about satisfaction. If I can get one good deal or even a great deal that I find every time, I'm much more likely to be satisfied.

I was just thinking about this. Costco either loses money on or just barely breaks even on their hotdogs, but they keep selling them at $1.50. It's a part of their brand. It would be smarter for them to raise the price of their membership another $10/yr to offset the losses than it would to raise the price of their hot dogs another $0.50 to make them profitable.

I wouldn't be surprised if costco writes off the loss on hot dogs as a marketing expense.

Re: The game theory of how algorithms can drive up prices

#118

Post by pessimizer »

parineum wrote: post #64

"Algorithm" has got to be the least useful word in English today. It isn't the software that's responsible driving up prices, it's the information. > Yet a widely cited 2019 paper (opens a new tab) showed that algorithms could learn to collude tacitly, even when they weren’t programmed to do so. A team of researchers pitted two copies of a simple learning algorithm against each other in a simulated market, then let t…

> The researchers quoted in the article are, essentially, defining collusion as knowing what competitors prices are.

It really is just another display about how injured the concept of "collusion" always was. Plenty of competitors have met in smokey rooms in order to fix prices, but you don't actually have to speak to each other to agree that if all of you can maximize your profits together, you should. Everybody knows how much everybody else is charging.

The ideal competition myth only works in a fictional zero-cost startup, zero-cost supplier, zero-cost distribution scenario. In real life if you try to enter a market with a few competitors with super high margins, they'll just threaten to freeze anybody who buys from you or sells to you out of the market, then offer you a ticket into their cartel.

Doesn't make it not evil, though, and it doesn't mean it's not an essential function of government to stop it. Government can't allow powerful little subgovernments to build up. You might as well allow paramilitary militias.

Re: The game theory of how algorithms can drive up prices

#119

Post by hollerith »

Workaccount2 wrote: post #116
Arwill wrote: post #91

Earlier quoted context omitted.

>see streaming companies increasing prices every few months They can do that because they are practically monopolies.

They can do it because people are hopelessly addicted to screens. You won't die if you stop watching Netflix. We aren't talking food or medicine here. In fact your life would probably improve. But addiction is a real animal.

I wish there were some term other than addiction here: addicts routinely steal from friends and family to feed their addiction; addicts who are parents sometimes threaten to stop allowing their children to visit with a grandparent unless the grandparent helps the addict pay for the addiction; drug addicts living in violent neighborhoods sometimes agree to murder somebody in exchange for drugs.

Screen addicts almost never stoop that low and the ones that do are addicted to a cam girl (e.g., Grant Amato), porn or gambling, not Netflix (or social media).

Re: The game theory of how algorithms can drive up prices

#120

Post by Workaccount2 »

friendzis wrote: post #8

> Imagine a town with two widget merchants. Customers prefer cheaper widgets, so the merchants must compete to set the lowest price. I always found this statement to be rather wishful. Individual lowering of prices makes sense if and only if your competitor is capable of saturating the market. Otherwise, demand elasticity becomes very relevant. Sure, your competitor may take the larger share of the market, but then y…

The AI market is a prime example of intense competition going on right now. All the dynamics are there.

If it was just one player, like Open AI, we'd still be at GPT-4 turbo and it would cost $400/mo.

Post reply on HN