Below is an excerpt from and link to a recent news article, “AI Flight Pricing Can Push Travelers to the Limit of Their Ability to Pay.” There have been several similar articles recently.
The paper describes the system as a “large market model” that uses an approach similar to generative AI systems, such as OpenAI’s Dall-E image generation tool. But instead of creating original art, the model is designed to ingest complex information about markets and to create pricing strategies that the paper says can dramatically raise revenue. Yerushalmi refers to this, rather unsubtly, as the “exploitation phase” … The blowback was quick. A group of Democrats in the Senate, led by Arizona’s Ruben Gallego, wrote a letter to Delta Chief Executive Officer Ed Bastian suggesting the program might “present data privacy concerns” and “will also likely mean fare price increases up to each individual consumer’s personal ‘pain point’ at a time when American families are already struggling with rising costs.”
I find this obnoxious.
It has me considering running a VPN continuously whenever I’m online. While doing so will not prevent companies with whom we have accounts, e.g., Amazon, from tracking and adjusting pricing to our “limit of our ability to pay,” it should help some.
Notwithstanding the small hit on battery life, does the hive mind recommend continuously running a VPN whenever online? Do you? Or am I merely spitting at the wind in doing so?
Dynamic Pricing… It is coming to us all over the place.
Buying from an iPhone? Sometimes a different price than buying from an Android device. Buying from an app? Sometimes a different price than buying from the website. Buying from a “poor” neighborhood? Buying from a “rich” neighborhood? Buying in the morning? At noon? In the evening? At night? Booking a hotel from Germany? Booking the same hotel from a different country? Booking a flight from the airline’s website? Booking a flight from a third party? You can go on about that. No end in sight. Even grocery stores start to think about dynamic pricing with those nice digital e-ink price labels that are entering or already have entered their brick-and-mortar shops.
I do not think that there is a real solution to that.
Regarding VPNs… VPNs do NOT prevent tracking. They prevent revealing your “true” IP address to the webserver you are visiting, and the traffic is encrypted (only) between you and the VPN server (which is a reason to consider a VPN if using public Wi-Fi access points). That is all that is being done.
So, the seller does not know from which IP address you are really buying when using a VPN. But, if the VPN’s IP address area is known, the seller knows that you are buying using a VPN (which is interesting at the same time). The seller still is able to track you when using a VPN via cookies, digital fingerprinting based on your device, and other things. And one important thing not to forget: the VPN provider is a third party that has access to a lot of data you are delivering: your true IP address, what you are doing on the web, and what not. So, it is very important to use a trusted VPN provider. Sometimes, delivering data to a VPN provider may be an issue on its own.
I am using a VPN connecting to my home (direct connection between my device and the router at home), when connecting to my NAS from the outside. Apart from that, I do not really use VPNs any longer: I have stopped using public Wi-Fi (with 4G/5G almost everywhere), and I have stopped watching “TV” in the US, Britain, or wherever else.
Regarding battery life: I have found that using a VPN is no big issue these days.
I am looking forward to different answers. This is an interesting topic.
Thanks for the thoughtful reply. I’m using ExpressVPN, which has a good reputation.
I was assuming that if I have my VPN set to a lower income state/area, that I might get better pricing. Is that not the case? Also, I could, with the corresponding friction, always browse in private mode (I nearly always use Safari) and on occasion when I need a Chromium-based browser, Brave.
Based on what you shared, it is probably a lost cause. Am I interpreting you correctly?
VPNs are notorious for tracking user actions and selling the data as their primary source of profits.
So your solution, in some ways, is worse than the problem.
TV’s, as briefly mentioned above, also sell a tremendous amount of tracking data.
Modern TVs have been described as consumer data harvesters with a free screen attached!
FYI, disconnecting a TV from the Internet and only manually updating its firmware, and then using a more consumer-friendly streaming box (AppleTV, for example) instead of Roku or other data harvesters can help, but there is no way to completely avoid being tracked as a tv or streaming user.
Nonetheless, if using a VPN does slightly reduce your online fingerprint and gives one peace of mind, by all means, continue to do so.
If the seller does not already know you or your device, it may or may not help you get better pricing, if they base their pricing on your IP address and if they do not know the IP address to be part of a VPN provider. If the seller is able to identify you or your true origin because of other data points or if the seller is basing the price on different data points (mobile, Apple device, time of day or whatever else), the VPN will have no effect.
Depending on what methods a seller is using in order to track, analyze, and/or identify the customer, changing/choosing browsers or browser modes can help. It is a cat and mouse game, though. Methods like fingerprinting (I currently have no good article on that at hand, so I am going with Device fingerprint - Wikipedia) can provide them with information no matter what. I am no forensic expert, so I am not 100% sure if there is a reliable way out of this misery. This is a reason why high-risk persons sometimes are using certain devices only for certain purposes in controlled environments (dynamic pricing for sure is no reason doing so).
Lost cause is a harsh description… If a different IP address is a factor for a different price at a certain online shop, it may have an effect. But there also can be a lot of other factors that may or may not have an effect on pricing. If you know the factors, you can try to check if changing a factor does change the price. An easy one is for instance to check an Amazon price from the Amazon app on your iPhone using the cell network and then to check the price from your desktop computer using a different network (not being logged in and with no active Amazon cookies on device). Different prices? In the past over here in Germany: yes, sometimes Amazon.de did have different prices (the one shown on the iPhone in the app was higher). I have no idea, if this still is happening.
EDIT (I am in no way affiliated to those creators):
P.S.
Regarding VPNs (very good article):
Or in podcast form:
With a very good (maybe even better) follow-up that details what a VPN actually does (and what VPNs or VPN services DO NOT do):
All those are not about VPNs and dynamic pricing, but not bad at all for learning more about VPNs. It bothers me that VPNs as such and VPN services are often talked about in the same way. However, they can mean two completely different things depending on how the term VPN is being used. Sorry for getting a bit off-topic here.
I can’t view the paywalled Bloomberg article but I’ve read about dynamic pricing recently. AI pricing looks a lot like what the grocery business once called zone pricing.
When I was 16 I started working for a major grocery chain and one of my jobs was stocking shelves. Which in those days included looking up the price of the item, in our zone, on a printed page and stamping the price on the can, carton, etc.
The page had prices for other zones in our state and the ones for our town were higher than for our much larger neighbor 18 miles away. Why? IMO, because they had a couple of large competitors and we didn’t.
I’ve not noticed differences in prices when using a VPN (ExpressVPN). But I have when I fail to clear my cache, like I normally do after each browser session
Back in the 1960’s our family went from traveling in a Buick to a VW Microbus. My father noticed that the cost of motel rooms dropped precipitously. So dynamic pricing has been around for a long time (actually, probably hundreds of not thousands of years) and does work both ways!
My mother had said (back in the 1940’s and 1950’s) that Sears had several grades of stores that got different qualities of merchandise depending on where they were located. Prices were the same for the different quality of merchandise. She noticed this when she moved to different areas.
Over a decade ago there were reports that Amazon.com was charging different customers different prices at the same time, but I think they quit the practice. Price trackers (Keepa, CamelCamelCamel, …) expose this immediately if they tried it again.
At one level the seller makes an offer to sell particular goods and services to you at a particular price, which you can choose to accept or decline, and that’s fair enough.
I agree with @Bmosbacker though, that “dynamic pricing” (whether using AI or not) which is designed to make an offer precisely at the “limit of our ability to pay” is obnoxious. I think I feel that, because of the imbalance of power between the corporation and the customer while trying to make a deal. They have many data points, including about me, on which to construct an offer. I have very limited insight into what those data points are, and even less into what the real costs to them are of providing the goods or services. It’s very hard to judge whether I am being asked to pay a fair or unfair price - and an unfair price where I am being exploited is basically a scam.
It’s yet another example of “enshittification”. A couple of decades ago I was part of a project to bring web access to a community facility in a small town in Tanzania. One of the unexpected benefits was learning, months later, that local farmers were using the web to check the daily market prices in the capital and were using that information to renegotiate the prices they’d accept from the middlemen who bought goods there for sale in the capital. It seemed to me exactly what the internet was for - information freely and easily available for all would give people better control of their lives. Sadly, the corporations and others have found many new and more sophisticated ways to subvert that, and for every improvement, there’s someone trying to make things worse for the rest of us.
Appearing to be in a different geographic location than you really are
Using a public WiFi in safer fashion. (Public WiFi isn’t safe ever, a VPN reduces the risk)
Upshot: I’m seriously considering switching to Brave
(Update: I replaced the links that didn’t work with screenshots that provide a high level summary)
I think the links do not show your results. They seem to show results for the browser I am using. Thanks for posting this. I was not aware of this site.
Hmm, I may consider doing something similar. I thought that there was some degree of preventing fingerprinting when using Safari. Am I wrong about that? I recall hearing something along those lines.
If we use a Mac/iPhone/iPad, we become a potential “high roller” for dynamic pricing, etc. It’s like wearing an expensive watch, it makes us stand out in the crowd. I’m not sure there is much we can do about this.
I turned it on for all browsing (it defaults to only working in private mode) and it dropped to 17.25 bits. I would imagine that it will improve as Apple works on it.
Gemini gave me a different take on randomised fingerprinting on iOS especially:
When you run a test on a site like coveryourtracks.eff.org, it’s testing for the raw data your browser provides. Since that raw data on iOS is largely a function of the underlying WebKit engine and your specific device, the fingerprint will almost always appear unique.
This doesn’t mean Brave or DuckDuckGo are failing. It means their most powerful protection on iOS is not randomization, but the proactive blocking of trackers that would use that unique data to build a profile of you. They prevent the trackers from ever seeing your fingerprint, which is an arguably more effective and direct form of protection than trying to fool them with randomized data.
The web server for whatever company you’re dealing with gets all of your information either way. So a sufficiently-motivated large company could put it together, whether or not the third-party trackers are being blocked.
No it isn’t correct. Safari is an easily fingerprinted browser. Further, while paying lip service to the idea, Apple’s efforts on fingerprinting are ~4yrs old. It’s a cat and mouse game. The mouse has learned new tricks and Apple is still watching the same old places.
@webwalrus I really wish I had faith that Apple would update their fingerprinting work. Unfortunately, I see no evidence to back my hope.
