I pulled the vendor's brochure; if you were curious what a vending machine would use this information for...
They collect demographics on WHO is purchasing what items including gender, age, etc. They use this information for targeting advertising (inc. with "partner media brokers").
They're also proud that when users install their app, it uses "gamification" to increase sales (whatever that means).
Vending machines covered by a large screen stand for everything I hate about contemporary tech.
They improve absolutely nothing from a buyer's perspective. Every step of the transaction is made worse. You can't glance at the entire inventory. You never know how much of an item is left. The machine does not reliably know how much of an item is left. Every interaction lags. And in return I get ads and mini games. Just so some C-suite cretin (guess what the C in C*O stands for) can show his little cretin friends how innovative his farts are.
It is late. I am hungry. My train departs in 2 minutes. Please, I just want a bloody Snickers.
> They improve absolutely nothing from a buyer's perspective.
It's the "good ads" argument -- this machine can guess your age/gender and offer up items most purchased by your demographic, all the while feeding that data back to the vendor. Who wouldn't love that? :-/
More seriously, this has been a standard capability in vending machines, fast food menu systems, and digital signage in general for over a decade. Check out this ad from 2012 for Intel's AIM Suite for an example of how this stuff is pitched: https://youtube.com/watch?v=KdMIp2vQjG8
"Is the viewer a teenage girl? Then change to content to highlight a back-to-school shoe sale a few doors down. Is it a senior male? Then why not tell him about the golf club sale at the sporting goods store?"
I'm fine with a machine taking my observable physical characteristics and using that to select ads for me. I'm also fine with ads that are not so much targeted as self-selected.
It's the use of non-observable, non-public info to build and reference an advertising profile where I start to object. It's creepy, and the ads are somehow less relevant.
The technique in restaurants is to track the number of items sold, add innovative special dishes every day or week, and at the end of every week or month, remove the worst selling items on the menu and replace it with one of the special dishes. This over time creates demand. Vending machines can do the same. Remove the worst selling item and replace it with anything. After a time, the vending machine will only sell items with demand for that location.
And this requires absolutely zero screens, cameras, or demographic data collection of any kind. Just track the inventory. This isn't new; I would be extremely surprised if Japanese vending machines didn't do just that for decades already.
But those cameras aren't there because of the vending machine. That is just a convenient platform. This is about ads, and tracking, and data brokering. We had digital ad screens placed on the platforms at train stations in the Netherlands with a camera hole a few years ago, and the advertising company swore that those cameras would not be activated, for now, until the legal side was resolved, or until they could get away with it without anyone noticing.
While I agree with you, the benefit is that it's automatic. That's the difference. It doesn't require anyone to count and track anything. It doesn't require thought or planning. It just happens. It doesn't require money spent on people to keep track of.
Whereas classic vending machine inventory systems requires an human person to track the re-fills or sales.
A happy medium would be no stupid cameras, but with electronic tracking and reporting of what is sold. That has to be a thing that can be done, right?
I wouldn't trust this automatic data for shit. When's the last time you used a vending machine that did not have any problems? The classic meme of people beating on a machine because something got stuck. How is that inventory managed? Does the inventory decrement every time someone pushes buttons to vend an item? Is it tracked by weight? Who enters the weight?
This seems like a system ripe for abuse by the manufacture on needing maintenance like the McDonald's soft server machine.
Your happy medium is what happens today in normal vending machines without cameras or giant screens. The normal boring glass-front ones that barely take credit cards.
The guy with a clipboard counting how many Diet Coke cans are missing hasn't been the way these things are managed since probably 2003.
That's exactly how the vending machines at my workplace operate. No screens or weird tracking (as long as you avoid the stupid "pay on your phone with our app" option). The machine knows exactly how many of what items it has dispensed and when without any of that nonsense.
My experience is the opposite - vending machines tend to run out of the really popular stuff and they just don't replace it until most of the crap stuff is gone.
Did you ever have the pleasure of encountering the US drugstore that replaced the entire set of doors of its entire refrigerated section with hulking, bright, animated, human-height-and-ultra-heavy TV screens?
The amount of effort that goes in to syncing up the contents of the fridge with what it is on the screen must be staggering. And not for a moment do I believe it reliably detects empty stock or the placement of the contents. Not to mention dealing with the extra heat it gives off.
Countless hours wasted and it is still worse than a glass panel in every way.
Not unlike orchestrating a dozen microservices and three Dooms worth of code to render a simple form to the three users a typical application has.
To Avakian, it’s simply an expected growing pain. Cooler Screens plans to educate customers about the digital displays and launch features like voice recognition, so shoppers can ask about prices or item locations.
“This is the future of retail and shopping,” Avakian said.
If this is our future of "education", these parasitic, data grubbing companies can fuck right off along with any larger company that supports them.
Nothing whatsoever about this in the least bit improves the time you spend grabbing something from a cooler. It's blatant enshitification of a very basic buying task that need not involve any further interface, and it so obviously does nothing but try to squeeze more money from said basic interaction for nobody's benefit except Walgreens, the companies paying for their grubby little digital visuals, and the scum-loaded company that's selling this garbage to them.
I can almost imagine a future in which even the tissue of toilet paper comes covered in "engagement" ad visuals impregnated into its material through some sort of micro-display technology. At last in that case it would get exactly the engagement it most deserves and afterwards go where it really belongs.
A little bit of an aside, but the absolute gol CNN website is impressive:
> We apologize, but your web browser is configured in such a way that it is preventing this site from implementing required components that protect your privacy and allow you to view and change your privacy settings. This functionality is required for privacy legislation in your region.
> We recommend you use a different browser or disable the “EasyList Cookie” filter from your “Content Filtering” settings (found under “Settings” -> “Shields” in the Brave Browser).
It turns out that whole thing was a grift where the former CEO left and founded the startup and then “sold” the Cooler Screens to his friends back at Walgreens. The new CEO came in and realized how dumb and expensive the screens were, killed the deal and is now being sued by the grifter.
The CEO at the end of the video talking about "educating" the consumers about the benefits of this new system really drives the point home that a large part of current tech is made and paid for by douchebags.
Yes, though at the time I believe it was static (no ads) so I was just left wondering why they felt the need to put up a TV screen showing what I'd see through the glass. It may have obscured whether something was in stock, but I forget.
The one upside I would offer is that with some of them, you can "click" the item to view nutritional facts and ingredients. That's invaluable to me personally, as frequently I want the "least sugary" option available in the machine.
That's quite the brochure. My favorite selling point is the way you can make the machine display products that it's doesn't actually have inside, sell them, and give the punter a digital IOU instead of the soda they tried to buy...
> "Customizable UI design - Product selection can be extended to include products not physically in the machine. Consumers can store them in Invenda Wallet [a mobile app] and redeem them somewhere else."
Somehow managing to turn a convenience-driven impulse buy into an additional chore to redeem later.
At the end of the day, there must be some angle I'm not understanding or these features wouldn't actually drive sales. I wonder if the idea is to vend digital products? Drive traffic to nearby physical stores through some kind of targeted digital coupon? Has anyone seen this kind of thing in the wild?
Wow... the whole point of a vending machine is to get a soda when you want to have a soda, if you wanted to have a soda later you could get it (usually a lot cheaper) at a supermarket? I hope there's at least a warning message where you can abort your purchase before you get the "voucher" instead of the actual product?
Hah! "Dude, you have a lot of NFTs in your Metamask wallet, what is going on?"
"Oh, those are just all the snacks I badly wanted but were out of stock in the vending machine and I can redeem them at some vague time at some vague location. Why, did you want to buy them?"
>> "Customizable UI design - Product selection can be extended to include products not physically in the machine. Consumers can store them in Invenda Wallet [a mobile app] and redeem them somewhere else."
Speaking only for myself, this would not be long-term profitable for the vending machine company, as the cost of them fixing the boot-sized hole in the machine would be far higher than the $2 they stole from me.
Place it somewhere where there are enough people that haven't tried it before and trick them into buying something they don't want. No one will ever use it more than once.
You're getting a soda's worth of money from every person who is tricked, and some will notice the trick before it happens or hear about it. I'm guessing it's not worth the price of the machine, tech, install costs and opportunity costs of a normal vending machine.
Detecting "gender" by facial recognition on a 21st century uni campus in the USA ... what could _possibly_ go wrong and cause a massive media meltdown?
The machines use the charitably named "demographic sensor" which is obviously the embedded camera linked to a "facial recognition" application, BUT, it doesn't appear that it actually recognizes (or records) faces. Meaning, it's not linking your face to any online identity, or recording your face at all. In fact, the company is European and claims that their entire platform is GDPR compliant, which is... probably true?
Rather, it throws out a series of guesses and confidence values of a person's age, gender, and race, and allow the planogram (and OOH advertising) to change dynamically based on that information.
Which is not necessarily great, but also an entire order of magnitude less invasive than every interaction any casual user has on the internet with any ad ever. Or, frankly, with any POS system recording a repeat purchase from your credit card, from which motivated vendors can back into the rest of the demographic data.
I'm not excusing it, but while the error reads "facial recognition", it's more "stereotyping enablement platform," which, while only marginally better, is probably still better.
It's also hilarious to think of the thing displaying the green M&M if you're a "probable" woman, and the red or yellow M&M if you're a "probable" man, and seeing how long it'd take for anyone to correlate the change.
That's what the company claims, and maybe it's right, but how do we know? Do we have access to the source code? Will it change in future? You can be absolutely sure that if the company thought they could make more money by storing the photographs or somehow associating them with real identities then they would.
In theory they can do something like "Why do you look so sad? Maybe a Mars can cheer you up!". This can be done while being GDPR complient but people don't like that. It creates a feeling that everything is watching you.
Privacy is not about hiding things, it is about controlling what you want to show.
Kind of surprised people haven't started putting a piece of tape over it, then coming back 8 hours later with a small drill and drilling through the now-blacked-out camera lens to permanently disable it in their living areas.
> In fact, the company is European and claims that their entire platform is GDPR compliant, which is... probably true?
I don't see how it is. The mere processing of my personal information (age, gender, race – doesn't matter that it's only a guess/estimate) without my explicit consent or any contractual need should already represent a violation of GDPR.
As far as I understand it's not, because GDPR concerns itself with personally identifiable data and "age/gender/race" is not identifiable in general (in context of a vending machine in a large city).
Correlate this with the date and location of the vending machine and it's not so clear anymore. GDPR is also concerned with data that could potentially be used to identify you.
> In fact, the company is European and claims that their entire platform is GDPR compliant, which is... probably true?
Is a picture of my face necessary to sell me a chocolate bar ? I would say not, and if they take it without asking consent before I would say that it is not GDPR compliant.
Honestly the whole thing would be funny if it's wasn't part of the dystopian society we're building. Is this... innovation now? What are we trying to optimize society for? Extracting maximum rents? But were will the money come from if everyone is doing the same thing and producing nothing of value?
Honestly, as a European some American racist tropes are truly baffling to me. Stereotypically Black people apparently enjoy watermelons, deep fried chicken and grape soda. In other words Black people allegedly like food humans often like and Americans managed to take than and make it racist.
I'm not saying it's not used as a racist trope. I'm saying it's amazing how Americans managed to be racist enough to make such a non-issue racist.
The racism is in the context of deprivation. Watermelon in and of itself is neutral. In America however, freed slaves did not have access to good agricultural land that could grow many fruits and vegetables. Watermelon grows on poor quality soil, so it was available to the poorest black sharecropper.
Because of this, it became associated with black people, and thereby with criminality, low education and poverty among the dominant culture that created this sort of pattern-matching.
My experience has been that european cultures are approximately as racist as the US, they just have different histories and dynamics so it comes out differently. There is a lot less visible reckoning with race as well, which makes it easier to assume or pretend that there are no problems with it.
Meet-and-greet the first few days at my college dorm nearly 20 years ago:
> Albino student: "Hey, where'd you get that watermelon?"
> Student with watermelon: "Dude, I'm black."
I think it was around then I first learned of the watermelon / friend chicken stereotype (I didn't get why everyone laughed when he said this), but "grape soda" is a new one to me.
>I'm saying it's amazing how Americans managed to be racist enough to make such a non-issue racist.
And you're a European? As a fellow European (balkans) I can comfortably say that many European cultures are racist as hell, to a degree that would make even quite a few Americans blush. I'd be careful about feeling any superiority over supposed European sophistication even in Western Europe.
Ask nearly anyone there about gypsies, jews or arabs in a context where they think they can speak honestly and privately. The kind of shit that gave Europe at least one huge holocaust and god knows how many pogroms before that hasn't entirely gone away, it's mostly just better camouflaged and with some of its sharper edges cut away.
Oh, absolutely. But the level of hyperfixation present in American racism against Black people stands out whereas the other racism feels more familiar.
I guess the European equivalent would be the racism against gypsies with the distinction being that it's less prominent for simple demographic reasons owing to us trying to genocide them for centuries. Of course this makes it all the more shocking to Americans when it does come up because of how flagrant it is.
But yeah, Europeans are no less racist than Americans. I'd go even so far as to say milquetoast non-racism (e.g. corporate DEI, the typical "one of each color" groups of people in stock images etc) is more widely accepted in the US than here in Europe given how many racist comments I see any time an ad includes a brown person. I think however that that has less to do with being less racist and more with being more "politically correct" (in the original sense -- i.e. saying what you're supposed to say, not what you believe).
Good. That at least will turn public opinion against this concept.
The whole notion of not even being allowed to see the same ads as your fellow man in a public space because you are involuntarily shunted into some kind of cohort is disgusting, regardless of what is being used to decide that.
Imagine walking somewhere and seeing all the digital ad screens around you switch to basket ball events or Kentucky Fried Chicken where they were showing tailor made suits just a second before. (Or ads for a pre-emptive rectal exam or mobility aids.)
I’m sure this is real but just for fun I’m entertaining the idea that someone is trolling — either with a fun dialog title or the app name.
I worked at a company in the late 2000s that moved to an office with automatic urinals. These were fairly novel at the time and had a matte black plastic unit with a flush handle that also had a shiny window made of dark, IR transparent plastic. It was clearly some kind of proximity sensor for the autoflush but some joker made an official “do not touch the cameras” sign that wound a few people up.
It’s probably something like Apple’s Face ID sensor, which is a “camera” in the strict sense of the word but it captures a depth map rather than an image. I’m willing to believe that they aren’t transmitting images or even raw depth maps over the Internet (for one thing, it would be a lot of data and it’s hard to conceive of what the use of such data would be), but they are almost certainly attempting to correlate demographic information, or possibly even individual users, with specific purchasing patterns.
They don't have to transmit images. It's enough to generate a unique hash that can track your activity in the presence of the panopticons. Sweeten the deal even more by capturing your IMSI for future reference.
In 2018 Renesas had a "Facial Expression Kit Giveaway" contest and I was one of the winners. It was the "RZ Omron Facial Expression Kit giveaway". RZ being one of their newer, at the time, Micro lines. I had to sign an affidavit to get my prize, which I did. Never did use it and it eventually went to a local Hamfest. Their example code was a vending machine.
Some places (colleges were very bad about this) would use your SSN as both an ID and a password - so to buy at the cafeteria you’d swipe your student ID and your PIN was the last four of your social.
Some workplaces did it too. It’s mostly gone now, but pieces still remain.
Given that many college students don’t even have SSNs (international students have no reason to get one until they want to work somewhere), this makes zero sense. What college does this?
Isn't facial recognition wide spread in the US yet?
In the UK we're already using it everywhere. Most stores and checkout systems have it. It's used in CCTV across the country. It's used by police to identify protestors. Schools in my area even install spyware on kids phones to monitor them and their families.
I can't imagine anyone would mind having being identified by a vending machine here lol... What's the risk?
When I read the story that schools in my area had been installing spyware on students phones I assumed it was some kind of isolated thing. But my GF works in schools so I asked her if she knew of this app and she was like, "oh yeah, all the kids have it".
I found this utterly insane but she seemed to think it was a good thing if it could be used to protect kids. And I guess I get the argument, but you could say the same thing about the government installing CCTV in everyone's homes... You need to draw lines, especially when there's an expectation of privacy.
This is my concern with facial recognition in shops too. If I go into a shop I expect people to see my face, so a simple CCTV system seems fine. I even expect people who know me might recognise who I am, so again, if a human being I've interacted with before identifies me on the CCTV system I'm okay with this. What I don't think people expect is to be identified by some automated national biometric system which checks to see if you've ever stolen anything before so it can alert security, while also trying to understand your emotions and individual purchasing habits.
I used to work for a large highstreet retail org and was literally pitched on a system which could identify our customers and tell us how they feel about their shopping experience and their shopping habits.
But like I say, no one cares. I've spoken about this with people so often and they act like I'm some kind of privacy lunatic. That said, people here get very animated if someone can access data they've personally shared to Facebook and made publicly available.
Some people choose to blindly trust the government and decide whatever it does is for the best and has no downsides. To quote someone I asked (unsuccessfully) for a referendum signature against a surveillance law, "If I can't trust the government anymore, then who can I trust?"
The risk is the countless little bits of info being gathered by all this tech falling into the hands of the next genocidal despot. I wonder which metric they'll base their cleansing around?
This stuff drives me mental, but the article ends up not being quite the black pill it could be. Nice to see students fighting back at the grassroots, the University (hopefully) acting, and people feeling enpowered to do something under the law.
We have to keep pushing for this stuff to become less and less normalised, and also for penalties to become more and more serious. When that happens, people will continue to feel more enpowered to fight this stuff. We should also make sure laws are clear that people have the right to "vandalise" such devices. In my opinion taking a bat to one of these things should at least be defensible in court.
I wonder where the line in the sand exists for what people feel is “okay” tracking and “not okay” tracking. Here it seems to be that almost universally, the students dislike the facial recognition. Many folks I’ve spoken to otherwise don’t care about being tracked online. I suppose people are more sensitive to being tracked when it transitions into the real world and becomes more “real.”
There’s an implied consent. I know that creating a Gmail account means that Google has access to all my emails. By using Gmail, I’m relatively aware that they’re building an ad profile on me, and I implicitly consent to that.
But if I found out they’re collecting the bloodwork pdfs my doctor sends and selling them to insurance companies, that would be objectively beyond any reasonable consent.
If you’re selling me M&Ms, I have a very very low tolerance for what I consider appropriate data collection.
You know, though I am not sure of how well-known it is outside of tech circles.
I agree with the second point. Facial recognition for demographic stats seems kinda reasonable if it's anonymised properly, but people don't like their photo being taken without knowledge of it.
I don't mind being targeted online since it changes the ads I see from "Acai berry smoothies" to "beach vacations". Something I am not interested in, to something I am interested in (due to browsing history). That makes my experience better and I understand the trade off while using a free product (Google Maps, Facebook etc etc).
However, the difference in this case is that the vending company (or someone along the chain) is using it to serve demographic ads to the customer of the vending machine. I still have to pay for the product and it's not changing non targeted ads to targeted ads (there were no ads before). I'm sure they aren't passing along some of the ad revenue to me since there's no competition (I'm sure all vending machines in the area are owned by the same company). So this is just a source of extra revenue for the company and serves no purpose to the user.
All right. I try my best to keep my adverts as random as possible(if they appear at all), since somehow it feels like I'm being passively influenced by viewing the same sort of adverts suggesting I buy stuff repeatedly.
The way they're sticking ads on everything with or without tracking is pretty annoying. I hope they lose money with the tech they had to dev and install vs ad returns.
I’m going to call bullshit on the claim that they don’t store data “It does not engage in storage, communication, or transmission of any imagery or personally identifiable information”
They obviously store data. You have to write the image to disk after the camera takes it. I guarantee that if a 3rd party audit were conducted they’d find data stored on there.
Heck, I’ll do the audit for zero upfront, payment contingent on finding images stored on there.
I know the HN guidelines say not to comment that someone hasn't read the article, but the comments section can be a little useless when literally all of the top comments I see are from posters who appear not to have read the article.
In summary (and, in fairness, these technical details are pretty far down the article):
> The machines are owned by MARS and the manufacturer is Invenda.
> MARS did not respond to requests for comment from CTV.
> Invenda also did not respond to CTV’s requests for comment but told Stanley in an email “the demographic detection software integrated into the smart vending machine operates entirely locally.”
> “It does not engage in storage, communication, or transmission of any imagery or personally identifiable information,” it continued.
> According to Invenda’s website, the Smart Vending Machines can detect the presence of a person, their estimated age and gender. The website said the “software conducts local processing of digital image maps derived from the USB optical sensor in real-time, without storing such data on permanent memory mediums or transmitting it over the Internet to the Cloud.”
> The website said the “software conducts local processing of digital image maps derived from the USB optical sensor in real-time, without storing such data on permanent memory mediums or transmitting it over the Internet to the Cloud.”
The most important thing to keep in mind about clauses like this is that at any time it's only one software update away from changing the behavior and suddenly starting to send everything. You'd never know. (And: That point in time may be in the past.)
What on Earth has ever given you the idea that I believe companies should be trusted in this regard?
I just quoted the most relevant technical details of what is actually presented in the article given that so much of the initial commentary when I first saw this post were from people who clearly hadn't read the article.
> “It does not engage in storage, communication, or transmission of any imagery or personally identifiable information,”
Probably bullshit.
Firstly, "personally identifiable information" is usually defined very specifically, like 'we didn't record their social security number'.
Secondly, they send a bunch of information extracted from the image and they probably store the raw images on the machines which are then manually extracted at some point. Like photocopier that store a scan of everything put on the bed in its own hard-drive.
This should be an obvious truth. But people who install secret facial recognition in vending machines are also likely to be the type of people would exclude face images from their definition of PII.
> Guess you’ve never heard of names? Or date of birth? Personal doesn’t mean unique.
Your parent commenter is contesting the "A face is a unique identifier" part. They are not contesting "A face is personal". Of course face should be PII but at the same time it is not a unique identifier, so your parent commenter is correct and you are correct too. But you are replying to an objection that was never made.
Thank you for the excerpt. I attempted to read the article but have a firmly negative reaction to the auto-playing video that CTV chose to pop over it.
I hate that with time there is just going to be more and more covert non-consentual facial recognition everywhere. It should be outlawed, but obviously that does not align with the interests of the powerful so it'll never happen.
there's a lot of roads all leading to the same place, there needs to be a constitutional amendment about the right to own your own data, and if someone collects it from you they need to pay you to collect it, and you should receive a tax on any transaction where its sold
at least that's my suggestion
if people collected my information and sent me a check in the mail every month for a decent amount, i might not be so annoyed.
Just smash it if you see it. Or someone suggested super glue. If we all actively destroy these devices, it'll eventually not be worth it to replace them.
Am a privacy advocate, yet think this might be an (understandable) overreaction.
I remember older generation tech used in digital signage applications that wanted to identify ages and genders to show them appropriate content, say apparel for example. It ran locally and was well meaning (enough).
Unfortunately in a world with Meta, ubiquitous telemetry, Ring doorbells, NSA, and data breaches and brokers, … we can’t trust anyone to do the right thing any longer.
So this type of benign-ish functionality, which would have been all local just two decades ago, unfortunately now can’t be trusted either.
Still, I’m very happy to hear a young person or two being concerned. Last trip to a college campus there was mandatory app and no cash accepted in many places.
I think the key difference here is that the digital signage technology you speak of, the recognition was a necessary requirement to show the appropriate content, clothing for example.
In this case, the recognition doesn't do anything for the end user of the vending machine. It is purely data for other brokers to use. The vending machine didn't say to the user that people of their demographic were buying Diet Coke over normal Coke.
Slapping facial recognition on something that didn't have it before, without any obvious functionality or benefit for the persons face being scanned is very much cause for concern, in my opinion.
This was an M&M (candy) machine. To really know, someone would need to audit the hardware and software, so we're only speculating. Can't definitively say without more info.
Was trying to make the point that there is a potential justification for the functionality, though tenuous.
I don't know why you think it's relevant. All that matters is the privacy law in the local jurisdiction. Some people will be fine with it, others wont. The real question is: what's the harm?
Because I don't necessarily think that local privacy law is all that great in the digital age, nor is our ability to ensure that we're complying with it, especially when third parties are involved.
People were clearly concerned about it, because the nature of its services weren't all that clear to the end user. Students protested, it was investigated and found that no PII information to taken, great no further action required.
My point is that the situation was clearly cause for concern because we have a lack of trust in these technologies and those who use them.
Case in point from the article "the embedded cameras inside Cadillac Fairview’s digital information kiosks used facial recognition technology to record over five million images of shoppers at malls without their customers’ knowledge or consent."
Exactly. All the company gets is probably anonymized aggregate data like "Product X is bought by men 40% and women 60%", which would pass the GDPR and any privacy protection laws. The point is that we don't have access to the software so we have to trust them about doing (only) that kind of aggregations. Furthermore, as we don't trust companies anymore, we fear that cameras are used for other not innocuous purposes.
By the way, does any 3rd party inspect the software of all those machines with cameras, before they are allowed to be installed? Probably not so it's their word against the suspicions of privacy wise people, until something goes wrong and some kind of investigation starts.
In that case I want the Sensor™ to be physically unable to send out actual images to the computer that handles the vending machine, so no software update can actually start doing something else than they claim it to do originally.
I'm pretty sure many big box places like Home Depot and Walmart are using actual facial recognition on everyone who is in or near any of their stores. This vending machine sounds pretty benign by comparison.
That is to say, my comment was meant to inform the concerned reader about other dastardly actors, not to excuse the behavior of the vending machine. (Not sure I have the authority even to make such an excuse.)
But you phrased it in such a way that downplayed the vending machine doing it, the three reaction you got, which implies there's a different wording that would have gotten your point across better.
Perhaps, but one of the rules of this forum is to apply the most charitable reading of comments, so I don't always try that hard to craft perfect responses.
Glob of epoxy putty over the camera. Simple fix. Everyone's bloviating about how "concerned" they are, but nobody can even think of a simple step to protect themselves? This is a university with smart students in it?
Making noise about this isn't a sign that students weren't smart enough to think to cover the camera. The article said that students had covered it.
But that's only a very immediate reactive response. A more proactive and impactful response would be to hurt the companies involved (e.g., bad PR, loss of sales, machine recall costs, legal problems), put fear into other companies doing things like that, and hopefully also prompt long-lacking better privacy legislation.
Those are perfectly effective, they're just easily removed. Something permanent like epoxy would be crossing the line into vandalism. I imagine nobody wants to take one for the team, seeing as how they're on camera.
That'd easily avoided by first putting a sticker over the hole and only later injecting some epoxy through the sticker. Do it neatly and they could even remove the sticker once the epoxy has hardened and paint over the former hole in a similar colour as the machine.
Putting a sticker on the camera inconveniences the college, who have to send someone to remove it. They are an innocent third party with no power to replace the vending machine with one without a camera.
Putting epoxy on the camera inconveniences the vending machine owner, who has to replace it. They're the assholes who put the camera there in the first place, and when they replace the machine they can easily put in one with no camera.
If you're going to engage in civil disobedience, why not do it right?
>>Putting a sticker on the camera inconveniences the college, who have to send someone to remove it.
The college doesn't care(or even know) about the facial recognition functionality, so presumably they don't care about this sticker either. If the sticker stops the machine from working, then the manufacturer would be called to fix it.
As multimodal AI becomes more prevalent and commercial AI attendants become the dominant user interface, there's going to be a basic expectation of object, gesture and facial recognition.
"Welcome back to McDonald's, Mr. Beattie, would you like your usual?" Nods head and grunts while sending text "Very good. I'll charge your card on record and send a notification when your meal is ready. Thank you!"
This will all require an integrated camera and microphone, and maybe other sensors as well. All turned on 24/7. This isn't science fiction, it's just a few years away from every day reality.
They collect demographics on WHO is purchasing what items including gender, age, etc. They use this information for targeting advertising (inc. with "partner media brokers").
They're also proud that when users install their app, it uses "gamification" to increase sales (whatever that means).
See here, they're super proud of it too:
https://a.storyblok.com/f/184550/x/e7435c019e/brochure-svm_g...